A powerful GitHub Code & Repo Analysis CLI for Researchers, designed to track activities, scrape and parse repository contexts, and search intelligently—without leaving YOUR terminal.
- GhResearcher 🔬
GhResearcher is a terminal-based toolkit built for researchers, developers, and tech enthusiasts who need to keep a close eye on the open-source ecosystem. Instead of navigating through web interfaces, GhResearcher leverages the GitHub REST API and GitHub CLI (gh) to bring timelines, repository structures, and multi-domain searches directly to your command line.
Whether you want to track a specific expert's "feed", monitor the activities of a lab/organization, or quickly dump a repository's structure into an LLM-friendly context file, GhResearcher provides the necessary utilities in a clean, terminal-native format powered by Rich and Typer.
- Curing "Free-Range" Research: Many researchers and grad students (especially in computational fields) feel isolated without daily guidance. This tool acts as your academic "social feed" for code. By tracking what experts and labs are actively working on, it keeps you involved, gives you clear targets, and ensures you stay motivated and closely aligned with mainstream developments.
- Terminal First: Keep you in the flow. No context switching to a web browser.
- Data Density: Present maximum information with minimal clutter. Long commit hashes are truncated; pagination is handled automatically.
- LLM-Friendly: Commands like
parseare explicitly designed to generate.mdfiles that can be directly fed into Language Models for code analysis and project understanding. - Privacy & Security: Relies entirely on your local
ghauthentication. No third-party servers, no telemetry.
- Dynamic Tracking (
monitor):- Track events from Users, Organizations (
--org), or specific Repositories (--repo). - Read a user's GitHub Feed (
--received) to see what experts are paying attention to. - Seamless pagination handling to bypass standard API constraints.
- Track events from Users, Organizations (
- Repository Parser (
parse):- Default mode exports README + repository tree into a Markdown file(similar to tree command).
--viewopens a pager instead of writing files.--view-mode readmeusesgh repo viewdirectly for README-only paging.--view-mode treeshows only the repository tree.--sourcelists savedAgent Parser(like DeepWiki OR some GitHub LLM-based parsers)source URLs for the repository.--sources-file <file>loads extra or overridden savedAgent Parsersource URLs from JSON.- Single-file targets like
owner/repo/path/to/filefetch file contents directly without cloning. - Supports
direct download of file contentin addition to paginated viewing.
- Multi-domain Search (
search):- Quickly search across repos, code, issues, and pull requests.
- Python 3.8+
- GitHub CLI (
gh): You must have the GitHub CLI installed and authenticated.# 1. Install gh (Ubuntu/Debian example) # follow https://github.com/cli/cli/blob/trunk/docs/install_linux.md#debian (type -p wget >/dev/null || (sudo apt update && sudo apt install wget -y)) \ && sudo mkdir -p -m 755 /etc/apt/keyrings \ && out=$(mktemp) && wget -nv -O$out https://cli.github.com/packages/githubcli-archive-keyring.gpg \ && cat $out | sudo tee /etc/apt/keyrings/githubcli-archive-keyring.gpg > /dev/null \ && sudo chmod go+r /etc/apt/keyrings/githubcli-archive-keyring.gpg \ && sudo mkdir -p -m 755 /etc/apt/sources.list.d \ && echo "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/githubcli-archive-keyring.gpg] https://cli.github.com/packages stable main" | sudo tee /etc/apt/sources.list.d/github-cli.list > /dev/null \ && sudo apt update \ && sudo apt install gh -y # 2. Authenticate with your GitHub account gh auth login
We recommend installing GhResearcher via pip for simplicity:
pip install ghresearcherOr clone the repository and install it via pip:
git clone https://github.com/MaybeBio/GhResearcher.git
cd GhResearcher
pip install -e .Verify the installation:
ghresearcher --helpThe monitor command provides a unified chronological timeline of GitHub activities.
Three important commands daily for me:
# you can change the target file to your own file
# yesterday to today activities of my followed experts
ghresearcher monitor -f /data2/GhResearcher/tests/target_user.txt --since $(date -d "1 day ago" +%Y-%m-%d) --expand-commitsUsage: ghresearcher monitor [OPTIONS] [TARGET]
❯ ghresearcher monitor --help
Usage: ghresearcher monitor [OPTIONS] [TARGET_NAME]
Track and view the recent events of specific GitHub user(s), organization(s), or repos. Supports single target directly or batch targets via file. Events from
multiple targets are combined into a global chronological timeline.
╭─ Arguments ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ target_name [TARGET_NAME] The GitHub user, org, or repo to monitor │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --file -f TEXT File containing GitHub targets (one per line) │
│ --received -r Fetch feed instead of user's own events │
│ --org -O Treat target as an Organization instead of a User │
│ --repo -R Treat target as a Repository (owner/repo format) │
│ --limit -l INTEGER Number of events to fetch per target [default: 30] │
│ --since TEXT Filter events on or after this date (YYYY-MM-DD) │
│ --until TEXT Filter events on or before this date (YYYY-MM-DD) │
│ --expand-commits Make additional API calls to get commit details for PushEvents if missing │
│ --help Show this message and exit. │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯Track the public actions of a specific developer (e.g., pushes, stars, forks).
# Monitor my activities (single user)
ghresearcher monitor MaybeBio -l 5
# Monitor multiple users (file input)
ghresearcher monitor -f users_to_track.txt
# ghresearcher monitor -f ./tests/target_user.txt --since 2026-05-11 --expand-commitswhat you get is:
Fetching events for target(s): MaybeBio...
2026-05-12 23:12:28 | ⭐️ MaybeBio starred alchaincyf/huashu-design
2026-05-12 20:07:14 | 🚀 MaybeBio pushed to MaybeBio/GhResearcher (no commit info)
2026-05-12 19:10:24 | 🚀 MaybeBio pushed to MaybeBio/bioinfor_script_modules (no commit info)
2026-05-12 19:02:18 | 🚀 MaybeBio pushed to MaybeBio/GhResearcher (no commit info)
2026-05-12 18:59:39 | 🆕 MaybeBio created branch 'main' at MaybeBio/GhResearcherif multiple, you get is
Fetching events for target(s): alexholehouse, Junjie-Zhu, HFChenLab, sirius777coder, Zuricho, Immortals-33, lujiarui, ChenDdon, AspirinCode, bjing2016, tyang816, prokia, Gonglab-THU...
2026-05-13 08:50:23 | ⭐️ prokia starred yliust/Tactile
2026-05-13 03:04:54 | ⭐️ Zuricho starred jsdoc/jsdoc
2026-05-12 23:14:59 | 🚀 AspirinCode pushed to AspirinCode/awesome-AI4MolConformation-MD
- [cc88d0d] (expanded) Update README.md
2026-05-12 22:54:36 | 🚀 AspirinCode pushed to AspirinCode/papers-for-molecular-design-using-DL
- [c062c62] (expanded) Update README.md
2026-05-12 20:29:29 | ⭐️ sirius777coder starred aqlaboratory/genie3
2026-05-12 20:01:38 | ⭐️ Immortals-33 starred obra/superpowers
2026-05-12 19:54:51 | ⭐️ Zuricho starred nicobailon/visual-explainer
2026-05-12 19:48:18 | ⭐️ Zuricho starred RomeroLab/BioDesignBench
2026-05-12 19:17:07 | 🔹 bjing2016 performed MemberEvent on MihirBafna/boltzgen
2026-05-12 19:08:38 | 🚀 AspirinCode pushed to AspirinCode/awesome-AI4MolConformation-MD
- [042c4c9] (expanded) Update README.md
2026-05-12 18:54:19 | 🚀 AspirinCode pushed to AspirinCode/awesome-AI4MolConformation-MD
- [6a0bac1] (expanded) Update README.md
2026-05-12 18:50:17 | 🚀 AspirinCode pushed to AspirinCode/awesome-AI4MolConformation-MD
- [348fb9f] (expanded) Update README.md
2026-05-12 18:24:54 | ⭐️ AspirinCode starred HealthRex/PhysicianBench
2026-05-12 16:23:29 | ⭐️ Zuricho starred yliust/Tactile
2026-05-12 16:19:36 | ⭐️ Zuricho starred smiles724/Proteo-R1
2026-05-12 09:56:57 | ⭐️ AspirinCode starred yaochenr/LLM-TPD-Extraction
2026-05-12 00:07:44 | 💬 tyang816 created issue 'Trouble recreating zero shot results on protein gym' in ai4protein/VenusREM
2026-05-11 23:30:54 | ⭐️ AspirinCode starred Yuan1z0825/nature-skills
2026-05-11 18:08:37 | 💬 Junjie-Zhu created issue 'Availability of pretrained weights' in Vincentx15/atomsurf
2026-05-11 17:51:13 | ⭐️ Immortals-33 starred openai/pluginsTrack the collective public events of an entire organization or lab. This is highly useful for following research groups' release updates or code pushes.
# Monitor GENTEL-lab's activities (single organization)
ghresearcher monitor GENTEL-lab --org
# Monitor multiple organizations (file input)
ghresearcher monitor -f labs_to_track.txt --org
# ghresearcher monitor -f ./tests/target_org.txt --since 2026-05-11 what you get is
Fetching events for target(s): GENTEL-lab...
2026-05-12 01:33:27 | ⭐️ justiniao starred GENTEL-lab/EVA
2026-05-11 10:59:50 | ⭐️ linjing-lab starred GENTEL-lab/VCWorld
2026-05-10 16:57:58 | ⭐️ NoahQue starred GENTEL-lab/VCWorld
2026-05-10 16:27:29 | ⭐️ Stern-612 starred GENTEL-lab/EnzymeCAGE
2026-05-09 21:31:27 | ⭐️ homandp starred GENTEL-lab/EVA
2026-05-09 15:54:25 | ⭐️ evacia starred GENTEL-lab/EVA
2026-05-09 08:40:08 | ⭐️ 7psusanpruitt40 starred GENTEL-lab/EVA
2026-05-09 00:21:22 | ⭐️ ardonw20 starred GENTEL-lab/EVA
2026-05-08 20:00:48 | ⭐️ CourageSiame starred GENTEL-lab/VCWorld
2026-05-08 16:44:58 | ⭐️ Aspectin starred GENTEL-lab/EnzymeCAGE
2026-05-08 08:24:37 | ⭐️ danserjeunjoline starred GENTEL-lab/EVA
2026-05-08 07:54:05 | ⭐️ inoue0426 starred GENTEL-lab/VCWorld
2026-05-08 07:11:21 | ⭐️ jpetchez starred GENTEL-lab/EVA
2026-05-08 06:25:01 | ⭐️ sb8587-a starred GENTEL-lab/EVA
2026-05-08 04:56:58 | ⭐️ evakagmadrit starred GENTEL-lab/EVA
2026-05-08 00:22:04 | ⭐️ Namkyeong starred GENTEL-lab/VCWorld
2026-05-07 21:13:59 | ⭐️ alexdebelka starred GENTEL-lab/VCWorld
2026-05-07 15:10:20 | 🐛 liwenqi-BGI opened issue in GENTEL-lab/EnzymeCAGE: 'FileNotFound'
2026-05-07 11:17:52 | 🍴 yishutu forked GENTEL-lab/GerNA-Bind
2026-05-06 14:46:47 | ⭐️ yifanfeng97 starred GENTEL-lab/EnzymeCAGE
2026-05-06 12:30:08 | ⭐️ bbyun28 starred GENTEL-lab/OriGene
2026-05-05 16:44:04 | ⭐️ heidban starred GENTEL-lab/EVA
2026-05-05 15:15:19 | ⭐️ FengxuSysbio starred GENTEL-lab/VCWorld
2026-05-05 15:06:26 | ⭐️ dingrenjie12 starred GENTEL-lab/EVA
2026-05-05 03:37:59 | ⭐️ stottlemartinsth starred GENTEL-lab/EVA
2026-05-05 02:58:17 | ⭐️ WaveoffBioMed starred GENTEL-lab/OriGene
2026-05-04 10:42:04 | ⭐️ shayeedew44d starred GENTEL-lab/EVA
2026-05-02 20:01:46 | ⭐️ ron5428-blantonl starred GENTEL-lab/EVA
2026-05-02 18:47:23 | ⭐️ william2014-jw starred GENTEL-lab/EVA
2026-05-02 18:09:32 | ⭐️ 81davejohnson80 starred GENTEL-lab/EVAif multiple, you get is
Fetching events for target(s): kiharalab, honig-lab, ai4protein, GENTEL-lab, steineggerlab, biomed-AI, BioComputingUP, ProteinDesignLab, sparks-lab-org, baker-laboratory, Graylab, isblab, THGLab, idptools, holehouse-lab,
Pappulab, KULL-Centre...
2026-05-13 09:37:56 | 💬 jmcavanagh created issue 'Getting invalid SMILES string while trying tutorial in T4 Colab' in THGLab/SmileyLlama
2026-05-13 06:29:17 | 🚀 jmcavanagh pushed to THGLab/SmileyLlama (no commit info)
2026-05-13 06:29:16 | 🔀 GbAlteri merged PR in THGLab/SmileyLlama
2026-05-13 06:10:58 | 🏷️ AntiMatter568 published release v1.0.1 in kiharalab/DAQplugin
2026-05-13 06:10:42 | 🚀 AntiMatter568 pushed to kiharalab/DAQplugin (no commit info)
2026-05-13 04:42:31 | 💬 MichaelChungyoun created issue 'Intuition of calculating the ProGen2 likelihood of the reverse sequence' in Graylab/FLAb
2026-05-13 04:42:31 | 🐛 MichaelChungyoun closed issue in Graylab/FLAb: 'Intuition of calculating the ProGen2 likelihood of the reverse sequence'
2026-05-13 04:36:09 | 🐛 MichaelChungyoun closed issue in Graylab/FLAb: 'missing data in immunogenicity data folder'
2026-05-13 04:36:08 | 💬 MichaelChungyoun created issue 'missing data in immunogenicity data folder' in Graylab/FLAb
2026-05-13 04:23:23 | 🔹 AntiMatter568 performed DeleteEvent on kiharalab/DAQplugin
2026-05-13 04:21:16 | 🚀 AntiMatter568 pushed to kiharalab/DAQplugin (no commit info)
2026-05-13 04:16:46 | 🏷️ AntiMatter568 published release v1.0.0 in kiharalab/DAQplugin
2026-05-13 04:16:19 | 🚀 AntiMatter568 pushed to kiharalab/DAQplugin (no commit info)
2026-05-13 03:55:16 | 🍴 BankBro forked biomed-AI/DiffDec
2026-05-13 03:46:31 | 🐛 MichaelChungyoun closed issue in Graylab/FLAb: ''tm' folder is missing'
2026-05-13 03:46:30 | 💬 MichaelChungyoun created issue ''tm' folder is missing' in Graylab/FLAb
2026-05-13 03:44:29 | 🔹 MichaelChungyoun performed DeleteEvent on Graylab/FLAb
2026-05-13 03:44:08 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:43:59 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:33:53 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:33:44 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:30:45 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:30:06 | 🆕 MichaelChungyoun created branch 'flab2-dev' at Graylab/FLAb
2026-05-13 03:27:22 | 🔹 MichaelChungyoun performed DeleteEvent on Graylab/FLAb
2026-05-13 03:26:54 | 🔹 MichaelChungyoun performed DeleteEvent on Graylab/FLAb
2026-05-13 03:24:26 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:22:54 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:20:52 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:20:36 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 03:06:54 | 🐛 MichaelChungyoun closed issue in Graylab/FLAb: 'When is the article updated'
2026-05-13 03:06:51 | 💬 MichaelChungyoun created issue 'When is the article updated' in Graylab/FLAb
2026-05-13 02:59:37 | 🐛 MichaelChungyoun closed issue in Graylab/FLAb: 'Rosace et al. binding data are likely in M'
2026-05-13 02:59:35 | 💬 MichaelChungyoun created issue 'Rosace et al. binding data are likely in M' in Graylab/FLAb
2026-05-13 02:59:11 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 02:53:53 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 02:51:54 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 02:38:15 | 🐛 MichaelChungyoun closed issue in Graylab/FLAb: 'HCDR3s swapped with trastuzumab HCDR1 in Shanehsazzadeh zero-shot dataset'
2026-05-13 02:38:13 | 💬 MichaelChungyoun created issue 'HCDR3s swapped with trastuzumab HCDR1 in Shanehsazzadeh zero-shot dataset' in Graylab/FLAb
2026-05-13 02:30:27 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 02:25:12 | 🚀 MichaelChungyoun pushed to Graylab/FLAb (no commit info)
2026-05-13 01:28:10 | ⭐️ alanfwilliams starred baker-laboratory/RoseTTAFold-All-Atom
2026-05-13 01:05:12 | 🚀 zlr-zmm pushed to ai4protein/VenusFactory2 (no commit info)
2026-05-13 01:02:35 | 🚀 zlr-zmm pushed to ai4protein/VenusFactory2 (no commit info)
2026-05-13 01:02:33 | 🔀 Patiskey merged PR in ai4protein/VenusFactory2
2026-05-13 00:40:24 | ⭐️ DSamuelHodge starred ProteinDesignLab/dEVA
2026-05-12 22:46:44 | 🐛 sudhir2016 opened issue in THGLab/SmileyLlama: 'Getting invalid SMILES string while trying tutorial in T4 Colab'
2026-05-12 22:26:19 | ⭐️ qianyhpku starred biomed-AI/DRlinker
2026-05-12 20:27:13 | 🔀 sooyoung-cha opened PR in steineggerlab/foldseek
2026-05-12 20:25:13 | 🚀 sooyoung-cha pushed to steineggerlab/foldseek (no commit info)
2026-05-12 20:14:03 | 🚀 sooyoung-cha pushed to steineggerlab/foldseek (no commit info)
2026-05-12 20:14:01 | 🔀 sooyoung-cha merged PR in steineggerlab/foldseek
2026-05-12 20:13:46 | 🔀 sooyoung-cha opened PR in steineggerlab/foldseek
2026-05-12 18:54:09 | ⭐️ zmzhang starred THGLab/SmileyLlama
2026-05-12 17:03:15 | ⭐️ zhimingzhang275 starred ai4protein/VenusX
2026-05-12 15:16:35 | ⭐️ Abbbbyyyy starred ai4protein/Pro-Prime
2026-05-12 14:40:01 | ⭐️ DrDiscoDao starred THGLab/HiQBind
2026-05-12 12:58:09 | ⭐️ insilicoscientist starred baker-laboratory/RoseTTAFold-All-Atom
2026-05-12 11:39:18 | ⭐️ hajuchan starred steineggerlab/foldseek
2026-05-12 06:05:49 | ⭐️ lorcai starred steineggerlab/StrucTTY
2026-05-12 05:47:44 | 🚀 gterashi pushed to kiharalab/DAQplugin (no commit info)
2026-05-12 05:47:19 | 🚀 AntiMatter568 pushed to kiharalab/DAQplugin (no commit info)
2026-05-12 05:37:45 | 🚀 gterashi pushed to kiharalab/DAQplugin (no commit info)
2026-05-12 03:53:09 | 💬 dmoypal created issue 'Missing Alignment Visualizations in html Output' in steineggerlab/foldseek
2026-05-12 02:38:03 | 🚀 ryanemenecker pushed to idptools/goose (no commit info)
2026-05-12 02:35:14 | 🔀 Patiskey opened PR in ai4protein/VenusFactory2
2026-05-12 02:33:55 | 🚀 ryanemenecker pushed to idptools/goose (no commit info)
2026-05-12 02:31:48 | 🚀 ryanemenecker pushed to idptools/goose (no commit info)
2026-05-12 02:28:56 | 🚀 ryanemenecker pushed to idptools/goose (no commit info)
2026-05-12 02:27:42 | 🏷️ ryanemenecker published release v0.2.5.1 in idptools/goose
2026-05-12 02:14:56 | 🚀 ryanemenecker pushed to idptools/goose (no commit info)
2026-05-12 01:33:27 | ⭐️ justiniao starred GENTEL-lab/EVA
2026-05-12 00:56:26 | ⭐️ rujinlong starred steineggerlab/StrucTTY
2026-05-12 00:27:15 | ⭐️ damrane starred ProteinDesignLab/dEVA
2026-05-12 00:07:44 | 💬 tyang816 created issue 'Trouble recreating zero shot results on protein gym' in ai4protein/VenusREM
2026-05-11 23:36:06 | 🚀 adelbke pushed to BioComputingUP/nest-mongo-acl (no commit info)
2026-05-11 23:35:26 | 🚀 adelbke pushed to BioComputingUP/nest-mongo-acl (no commit info)
2026-05-11 23:01:28 | 🚀 LunaJang pushed to steineggerlab/StrucTTY (no commit info)
2026-05-11 18:42:35 | 🚀 LunaJang pushed to steineggerlab/StrucTTY (no commit info)
2026-05-11 18:21:16 | ⭐️ Hidroxiapatito starred steineggerlab/colabfold-protocol
2026-05-11 13:41:39 | 🔹 gamcil performed DeleteEvent on steineggerlab/foldmason
2026-05-11 13:41:35 | 🚀 gamcil pushed to steineggerlab/foldmason (no commit info)
2026-05-11 13:41:34 | 🔀 gamcil merged PR in steineggerlab/foldmason
2026-05-11 13:40:56 | 🔀 gamcil opened PR in steineggerlab/foldmason
2026-05-11 13:05:16 | 🚀 gamcil pushed to steineggerlab/foldmason (no commit info)
2026-05-11 10:59:50 | ⭐️ linjing-lab starred GENTEL-lab/VCWorld
2026-05-11 10:33:03 | ⭐️ samuelmcurtis starred baker-laboratory/RoseTTAFold-All-Atom
2026-05-11 10:19:35 | ⭐️ chengwilliamlin starred idptools/starling
2026-05-11 09:06:44 | ⭐️ AndyCycle starred steineggerlab/foldseek
2026-05-11 09:02:37 | 🚀 gterashi pushed to kiharalab/DAQplugin (no commit info)
2026-05-11 02:55:34 | ⭐️ wyqmath starred idptools/starlingFocus cleanly on the event stream of a single repository (e.g., releases, issues, PRs, stars).
# Monitor a single repository (isblab/disobind)
ghresearcher monitor isblab/disobind --repo -l 20
# Monitor multiple repositories (file input)
ghresearcher monitor -f repos_to_track.txt --repowhat you get is
Fetching events for target(s): isblab/disobind...
2026-05-01 18:18:10 | ⭐️ Raghav0573 starred isblab/disobind
2026-04-29 16:40:05 | 🍴 ipcamit forked isblab/disobind
2026-04-23 13:41:51 | 🍴 paolellopotanovic-ctrlxiaoke forked isblab/disobindDiscover new tools by looking at what an expert is watching. This fetches the "received events" feed (similar to the GitHub homepage feed).
ghresearcher monitor teorth --receivedwhat you get is
Fetching events for target(s): teorth...
2026-05-12 21:48:31 | 🍴 benediktjohannes forked teorth/estimates
2026-05-12 21:48:20 | 🍴 benediktjohannes forked teorth/equational_theories
2026-05-12 21:22:02 | 🍴 benediktjohannes forked AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 20:36:01 | 🔀 Milian0402 opened PR in teorth/erdos-guy-selfridge
2026-05-12 20:14:06 | ⭐️ jonahinthewhale starred AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 14:35:22 | 🍴 nachose forked teorth/erdos-guy-selfridge
2026-05-12 12:04:37 | 🔹 teorth performed GollumEvent on AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 12:00:33 | 🚀 teorth pushed to AlexKontorovich/PrimeNumberTheoremAnd (no commit info)
2026-05-12 12:00:33 | 🐛 teorth closed issue in AlexKontorovich/PrimeNumberTheoremAnd: '[CH2]: Limiting integral formula for smoothly
truncated Dirichlet series (Proposition 2.3)'
2026-05-12 12:00:31 | 🔀 anhhuyalex merged PR in AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 12:00:26 | 🔹 teorth performed PullRequestReviewEvent on AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 11:59:56 | 🔹 teorth performed GollumEvent on AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 11:40:39 | 🐛 github-actions assigned issue in AlexKontorovich/PrimeNumberTheoremAnd: '[BKLNW]: Uniform medium size bound on
theta (Corollary 8.1b)'
2026-05-12 11:40:23 | 💬 Yu-Misaka created issue '[BKLNW]: Uniform medium size bound on theta (Corollary 8.1b)' in
AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 09:17:18 | 🐛 github-actions assigned issue in AlexKontorovich/PrimeNumberTheoremAnd: '[CH2]: Limiting integral formula for
smoothly truncated Dirichlet series (Proposition 2.3)'
2026-05-12 09:17:14 | 💬 anhhuyalex created issue '[CH2]: Limiting integral formula for smoothly truncated Dirichlet series (Proposition
2.3)' in AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 09:17:10 | 💬 anhhuyalex created issue '[CH2]: Limiting integral formula for smoothly truncated Dirichlet series (Proposition
2.3)' in AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 09:15:31 | 🔀 anhhuyalex opened PR in AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 09:10:35 | 🐛 github-actions assigned issue in AlexKontorovich/PrimeNumberTheoremAnd: '[FKS2]: Converting bounds for theta into
bounds for pi (Theorem 6)'
2026-05-12 09:10:24 | 💬 Osalotioman created issue '[FKS2]: Converting bounds for theta into bounds for pi (Theorem 6)' in
AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 08:45:29 | 🐛 teorth closed issue in AlexKontorovich/PrimeNumberTheoremAnd: '[FKS2]: Matching lower bound on E_π (Theorem 6,
substep 2), cowritten with Grok'
2026-05-12 08:44:44 | 🚀 teorth pushed to AlexKontorovich/PrimeNumberTheoremAnd (no commit info)
2026-05-12 08:44:44 | 🐛 teorth closed issue in AlexKontorovich/PrimeNumberTheoremAnd: '[FKS2]: Upper bound on E_pi (Theorem 6, substep
1)'
2026-05-12 08:44:43 | 🔀 Osalotioman merged PR in AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 08:44:38 | 🔹 teorth performed PullRequestReviewEvent on AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 06:11:25 | 🐛 illdreamt opened issue in AlexKontorovich/PrimeNumberTheoremAnd: '[FKS2]: Matching lower bound on E_π (Theorem 6,
substep 2), cowritten with Grok'
2026-05-12 03:10:59 | ⭐️ orfyus starred teorth/symmetric_project
2026-05-12 02:14:48 | 🔀 Milian0402 opened PR in teorth/erdos-guy-selfridge
2026-05-12 00:31:13 | ⭐️ orfyus starred AlexKontorovich/PrimeNumberTheoremAnd
2026-05-12 00:00:59 | 💬 Osalotioman created issue '[FKS2]: Upper bound on E_pi (Theorem 6, substep 1)' in
AlexKontorovich/PrimeNumberTheoremAndPass a text file with multiple targets (one per line) to merge their activities into a single global timeline.
ghresearcher monitor -f experts.txt --since 2026-05-01 --until 2026-05-12By default, long pushes are truncated. Use --expand-commits to fetch detailed commit messages via additional API calls.
ghresearcher monitor teorth --expand-commitswhat you get is
Fetching events for target(s): teorth...
2026-05-13 08:57:32 | 🚀 teorth pushed to teorth/erdos-guy-selfridge
- [617f9ad] (expanded) Merge pull request #101 from Milian0402/maxiboi/readme-roadmap
2026-05-13 08:57:21 | 🔹 teorth performed PullRequestReviewEvent on teorth/erdos-guy-selfridge
2026-05-13 08:52:04 | 🚀 teorth pushed to teorth/erdos-guy-selfridge
- [0da591c] (expanded) Merge pull request #102 from Milian0402/maxiboi/c1-constant-docs
2026-05-13 08:51:57 | 🔹 teorth performed PullRequestReviewEvent on teorth/erdos-guy-selfridge
2026-05-13 06:17:43 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [bcfd513] (expanded) integral form of additivity
2026-05-13 05:37:12 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [fc850d0] (expanded) trim docstring
2026-05-13 05:35:07 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [a75fbf5] (expanded) Merge branch 'stieltjes' of https://github.com/leanprover-community/mathlib-at-ICERM26 into stieltjes
2026-05-13 02:10:07 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [92e6786] (expanded) redefine Stieltjes integral to handle backwards integral
2026-05-12 22:30:35 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [ad35844] (expanded) some map API
2026-05-12 20:20:08 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [2bce97a] (expanded) some simple lemmas about intervals
2026-05-12 16:57:47 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [c025070] (expanded) integration of constants
2026-05-12 15:38:00 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [3ffc66b] (expanded) remove warning
2026-05-12 15:35:29 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [e297d8a] (expanded) add connections to standard integrals
2026-05-12 15:28:46 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [92c1ea3] (expanded) notation for integral
2026-05-12 15:22:09 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [e7fbac7] (expanded) even more linearity API
2026-05-12 15:18:39 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [09e1ecb] (expanded) more linearity API
2026-05-12 15:06:36 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [bcfd5f1] (expanded) sectioning
2026-05-12 15:03:06 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [98ae4aa] (expanded) Stieltjes integral linearity API
2026-05-12 14:44:10 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [bedff9f] (expanded) automated style fixes
2026-05-12 14:38:46 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [0f90350] (expanded) bundle ofDiff as a hom
2026-05-12 13:59:52 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [18d419b] (expanded) BoxAdditiveMap API
2026-05-12 13:29:05 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [2a43bd9] (expanded) fix lean
2026-05-12 13:28:23 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [096a002] (expanded) fix lean
2026-05-12 13:27:31 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [3425dd1] (expanded) add other predicates for Stieltjes integrability
2026-05-12 13:16:48 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [6d9643a] (expanded) notational golf
2026-05-12 13:10:50 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [5198afa] (expanded) change interval to Ioc
2026-05-12 13:08:16 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [354ba9b] (expanded) generalize integration against summatory function
2026-05-12 13:00:43 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [44a86ee] (expanded) golf namespaces
2026-05-12 12:54:26 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [a692262] (expanded) add docstring
2026-05-12 12:49:38 | 🚀 teorth pushed to leanprover-community/mathlib-at-ICERM26
- [cfc0361] (expanded) change from Unit to Fin 1Parses a repository, file, or saved source catalog into terminal-friendly text. The default repository path uses GitHub's Trees API first, so it does not need to clone the repository unless the API response is truncated or unavailable.
Usage: ghresearcher parse [TARGET] [--view] [--view-mode readme|tree|both] [--source] [--sources-file FILE]
❯ ghresearcher parse --help
Usage: ghresearcher parse [OPTIONS] TARGET
Parse a repository, file, or source URL into Markdown/text.
╭─ Arguments ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ * target TEXT The GitHub repo (owner/repo) or file (owner/repo/path/to/file) [required] │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --output -o TEXT Output file path │
│ --view View in a pager instead of writing to disk │
│ --view-mode TEXT View mode for repositories: both, readme, or tree [default: both] │
│ --source List saved source URLs for the repository │
│ --sources-file TEXT JSON file containing extra or overridden source URLs │
│ --help Show this message and exit. │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
⚠️ Parse logic is global, i.e., not adding--viewoption will download the repository content to a file. Adding--viewoption will only view in a pager, without writing to disk.
# Write README + tree to a file
ghresearcher parse isblab/disobind -o Disobind_Context.md
# View README + tree in a pager
ghresearcher parse isblab/disobind --view
# README-only pager using the native `gh repo view` experience
ghresearcher parse isblab/disobind --view --view-mode readme
# Tree-only pager
ghresearcher parse isblab/disobind --view --view-mode tree
# Single file, e.g., README.md, you can directly append the file path after owner/repo
ghresearcher parse isblab/disobind/README.md --view
# Saved source catalog for the repository (we provide some default urls)
ghresearcher parse isblab/disobind --source --view
# Load extra sources from a JSON file (except default urls, you can also add your own urls in the JSON file, we will merge them together)
ghresearcher parse isblab/disobind --source --sources-file ./sources.json --viewCurrent default sources:
[
{"name": "deepwiki", "kind": "template", "url": "https://deepwiki.com/{owner}/{repo}"},
{"name": "zreadai", "kind": "template", "url": "https://zread.ai/{owner}/{repo}"},
{"name": "readmex", "kind": "template", "url": "https://readmex.com/{owner}/{repo}"},
{"name": "gitdiagram", "kind": "template", "url": "https://gitdiagram.com/{owner}/{repo}"},
]Implementation notes:
- Repository trees are fetched from the GitHub Trees API first.
- If the Trees API is truncated or fails for a very large repository, GhResearcher falls back to a temporary shallow clone to generate the tree.
--view --view-mode readmeusesgh repo viewdirectly, so it keeps GitHub CLI's native pager behavior.sources.jsoncan define both template sources and fixed URLs.- You can point
--sources-fileat any JSON file containing extra or overridden saved source URLs. - A ready-to-edit example lives at
sources.example.jsonin the repository root.
🌟 A Simple Example like:
I want to view the README of a repository, I first check the README to quickly understand the project.
ghresearcher parse Junjie-Zhu/IDPFold --view --view-mode readme
and then I want to quickly understand the file structure of the repository, so I can check where the main code lives, and where the data lives.
ghresearcher parse Junjie-Zhu/IDPFold --view --view-mode tree
e.g., I am pretty interested in script /src/common/pdb_utils.py, I can check it by running the following command:
ghresearcher parse Junjie-Zhu/IDPFold/src/common/pdb_utils.py --view 
finally, I want to check if there are any LLM Annotations for this script.
hresearcher parse Junjie-Zhu/IDPFold --source --view
I can directly open the default smart reading page in the browser to view the detailed information of the project.
Mine GitHub's powerful search engine directly from the terminal across five domains: repos, code, issues, PRs, and commits. GhResearcher fully wraps the underlying gh search and supports two interchangeable usage styles:
- Direct CLI flags: for quick one-off searches, with short flags for all high-frequency filters.
- Declarative YAML config: for saving frequent or complex search conditions and reusing them with one command.
When a CLI flag and a YAML field both appear, the CLI flag wins and overrides the corresponding YAML field.
ghresearcher search [OPTIONS] [item_type] [query]item_type: search type —repos/code/issues/prs/commits. Omit it if provided via the YAMLitem_typefield.query: search keywords.codesearch requires aquery; the other four (repos/issues/prs/commits) allow omittingqueryfor pure flag-based filtering, e.g.:
ghresearcher search repos -o microsoft --visibility publicFull help:
❯ ghresearcher search --help
Usage: ghresearcher search [OPTIONS] [item_type] [query]
Search GitHub across multi-domains (repos, code, issues, prs, commits) with full query support.
Accepts command line arguments, a structured YAML config profile, or a combination of both.
CLI flags always take precedence over YAML config values.
Examples:
ghresearcher search repos "LLM agent" -L Python -t artificial-intelligence -s stars -l 20
ghresearcher search code "TODO" -r MaybeBio/GhResearcher -f "*.py" -e py
ghresearcher search prs "fix bug" -o microsoft --merged
ghresearcher search --config examples/search_ai_repos.yaml
╭─ Arguments ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ item_type <str> Type to search: repos, code, issues, prs, commits │
│ query <str> The search query │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│ --config -c <path> Path to YAML search config file │
│ --web -w Open the search query in the web browser │
│ --json <str> Output JSON with the specified fields (comma-separated) │
│ --jq <str> Filter JSON output using a jq expression │
│ --template <str> Format JSON output using a Go template │
│ --limit -l <int> Maximum number of results │
│ --sort -s <str> Sort criteria │
│ --order -O <str> Order of results: asc|desc │
│ --owner -o <str> Filter on repository owner │
│ --repo -r <str> Filter on repository (owner/repo) │
│ --language -L <str> Filter by programming language │
│ --visibility <str> Filter by visibility: public|private|internal │
│ --match <str> Restrict search to specific field │
│ --topic -t <str> Filter on repository topic │
│ --license <str> Filter by license type │
│ --stars <str> Filter on number of stars (e.g. '>=100') │
│ --forks <str> Filter on number of forks (e.g. '>=10') │
│ --size <str> Filter on size range in KB (e.g. '5000..10000') │
│ --created <str> Filter on created date (e.g. '>=2023-01-01') │
│ --updated <str> Filter on last updated date │
│ --archived <str> Filter based on archived state: true|false │
│ --include-forks <str> Include forks: false|true|only │
│ --good-first-issues <str> Filter on number of 'good first issue' labels │
│ --help-wanted-issues <str> Filter on number of 'help wanted' labels │
│ --number-topics <str> Filter on number of topics │
│ --followers <str> Filter on number of followers │
│ --extension -e <str> Filter on file extension │
│ --filename -f <str> Filter on filename │
│ --label <str> Filter on label │
│ --state <str> Filter by state: open|closed │
│ --author <str> Filter by author │
│ --assignee <str> Filter by assignee │
│ --mentions <str> Filter by @mentions │
│ --milestone <str> Filter by milestone title │
│ --comments <str> Filter on number of comments (e.g. '>100') │
│ --no-assignee Filter on missing assignee │
│ --no-label Filter on missing label │
│ --no-milestone Filter on missing milestone │
│ --no-project Filter on missing project │
│ --include-prs Include pull requests in results │
│ --locked Filter on locked conversation status │
│ --closed <str> Filter on closed date (e.g. '>=2023-01-01') │
│ --interactions <str> Filter on number of reactions and comments │
│ --reactions <str> Filter on number of reactions │
│ --app <str> Filter by GitHub App author │
│ --commenter <str> Filter based on comments by user │
│ --involves <str> Filter based on involvement of user │
│ --project <str> Filter on project board (owner/number) │
│ --team-mentions <str> Filter based on team mentions │
│ --draft Filter based on draft state │
│ --merged Filter based on merged state │
│ --base -B <str> Filter on base branch name │
│ --head -H <str> Filter on head branch name │
│ --checks <str> Filter by check status: pending|success|failure │
│ --review <str> Filter by review status: none|required|approved|changes_requested │
│ --review-requested <str> Filter on requested reviewer │
│ --reviewed-by <str> Filter on user who reviewed │
│ --merged-at <str> Filter on merged date (e.g. '>=2023-01-01') │
│ --committer <str> Filter by committer │
│ --hash <str> Filter by commit hash │
│ --merge Filter on merge commits │
│ --author-date <str> Filter on authored date │
│ --author-email <str> Filter on author email │
│ --author-name <str> Filter on author name │
│ --committer-date <str> Filter on committed date │
│ --committer-email <str> Filter on committer email │
│ --committer-name <str> Filter on committer name │
│ --parent <str> Filter by parent hash │
│ --tree <str> Filter by tree hash │
│ --help Show this message and exit. │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯| Short | Long | Description |
|---|---|---|
-w |
--web |
Open the search results page in a browser |
| — | --json <fields> |
Output specified fields as JSON (comma-separated, e.g. "name,stargazersCount") |
| — | --jq <expr> |
Filter JSON output with a jq expression |
| — | --template <tpl> |
Format JSON output with a Go template |
| Short | Long | Description |
|---|---|---|
-l |
--limit <int> |
Maximum number of results (default 30) |
-s |
--sort <str> |
Sort criteria (varies by type, see §3.4; unsupported for code) |
-O |
--order <asc|desc> |
Sort order (only effective when --sort is specified) |
| Short | Long | Applies to | Description |
|---|---|---|---|
-o |
--owner <str> |
all | Restrict to a repo owner (org/user) |
-r |
--repo <owner/repo> |
code/issues/prs/commits | Restrict to a specific repository |
-L |
--language <str> |
repos/code/issues/prs | Filter by programming language |
| — | --visibility <str> |
repos/issues/prs/commits | Visibility: public/private/internal |
| — | --match <str> |
repos/code/issues/prs | Restrict the searched field (see per-type notes) |
| Short | Long | Description |
|---|---|---|
-t |
--topic <str> |
Filter by repository topic |
| — | --license <str> |
Filter by license (e.g. mit, apache-2.0) |
| — | --stars <num> |
Filter by star count (e.g. >=100) |
| — | --forks <num> |
Filter by fork count (e.g. >=10) |
| — | --size <range> |
Filter by size (KB, e.g. 5000..10000) |
| — | --created <date> |
Filter by creation date (e.g. >=2023-01-01) |
| — | --updated <date> |
Filter by update date |
| — | --archived <true|false> |
Filter by archived status |
| — | --include-forks <false|true|only> |
Whether to include fork repos |
| — | --good-first-issues <num> |
Filter by number of good-first-issues |
| — | --help-wanted-issues <num> |
Filter by number of help-wanted issues |
| — | --number-topics <num> |
Filter by number of topics |
| — | --followers <num> |
Filter by follower count |
| — | --match <name|description|readme> |
Restrict the searched field |
| Short | Long | Description |
|---|---|---|
-e |
--extension <str> |
Restrict to a file extension (e.g. py) |
-f |
--filename <str> |
Restrict to a filename (wildcards allowed, e.g. "*.py") |
| — | --match <file|path> |
file matches content, path matches path |
| — | --size <range> |
Filter by file size (KB) |
codesearch is powered by GitHub's legacy code search engine and does not support--sort/--order; if passed by mistake, GhResearcher warns and ignores them.
| Short | Long | Description |
|---|---|---|
| — | --label <str> |
Filter by label |
| — | --state <open|closed> |
Filter by state |
| — | --author <str> |
Filter by author |
| — | --assignee <str> |
Filter by assignee |
| — | --mentions <str> |
Filter by @mention |
| — | --milestone <str> |
Filter by milestone title |
| — | --comments <num> |
Filter by comment count (e.g. >100) |
| — | --no-assignee |
Only items with no assignee (presence flag) |
| — | --no-label |
Only items with no label (presence flag) |
| — | --no-milestone |
Only items with no milestone (presence flag) |
| — | --no-project |
Only items with no project (presence flag) |
| — | --locked |
Filter by locked status (presence flag) |
| — | --closed <date> |
Filter by close date |
| — | --created <date> |
Filter by creation date |
| — | --updated <date> |
Filter by update date |
| — | --interactions <num> |
Filter by interaction count (comments + reactions) |
| — | --reactions <num> |
Filter by reaction count |
| — | --app <str> |
Filter by GitHub App author |
| — | --commenter <str> |
Filter by commenter |
| — | --involves <str> |
Filter by involved user |
| — | --project <owner/number> |
Filter by project board |
| — | --team-mentions <str> |
Filter by team mention |
| — | --archived <true|false> |
Filter by archived status |
| — | --match <title|body|comments> |
Restrict the searched field |
| Short | Long | Description |
|---|---|---|
| — | --include-prs |
Include PRs in results (presence flag) |
| Short | Long | Description |
|---|---|---|
| — | --draft |
Filter by draft status (presence flag) |
| — | --merged |
Filter by merged status (presence flag) |
-B |
--base <str> |
Filter by base branch name |
-H |
--head <str> |
Filter by head branch name |
| — | --checks <pending|success|failure> |
Filter by check status |
| — | --review <none|required|approved|changes_requested> |
Filter by review status |
| — | --review-requested <str> |
Filter by requested reviewer |
| — | --reviewed-by <str> |
Filter by reviewer |
| — | --merged-at <date> |
Filter by merge date |
| Short | Long | Description |
|---|---|---|
| — | --author <str> |
Filter by author |
| — | --committer <str> |
Filter by committer |
| — | --hash <str> |
Filter by commit hash |
| — | --merge |
Only merge commits (presence flag) |
| — | --author-date <date> |
Filter by author date |
| — | --author-email <str> |
Filter by author email |
| — | --author-name <str> |
Filter by author name |
| — | --committer-date <date> |
Filter by committer date |
| — | --committer-email <str> |
Filter by committer email |
| — | --committer-name <str> |
Filter by committer name |
| — | --parent <str> |
Filter by parent commit hash |
| — | --tree <str> |
Filter by tree hash |
gh search has two kinds of boolean flags; GhResearcher converts each to the correct syntax automatically:
| Type | Example | CLI | YAML | Generated gh arg |
|---|---|---|---|---|
| Value boolean | --archived |
--archived true / --archived false |
archived: false |
--archived=true / --archived=false |
| Presence flag | --draft --merged --no-assignee etc. |
just --draft (omitted = off) |
draft: true appears / draft: false omitted |
--draft |
- Value boolean has only one member,
archived, and it must carrytrue/false. - Presence flags include:
web,draft,merged,include-prs,locked,merge,no-assignee,no-label,no-milestone,no-project. They only need to "appear" to mean true; a YAMLfalseis automatically omitted.
Valid --sort values depend on item_type; passing a wrong one triggers a warning:
| item_type | Valid values |
|---|---|
repos |
forks, help-wanted-issues, stars, updated |
issues |
comments, created, interactions, reactions, reactions-+1, reactions--1, reactions-heart, reactions-smile, reactions-tada, reactions-thinking_face, updated |
prs |
comments, reactions, reactions-+1, reactions--1, reactions-smile, reactions-thinking_face, reactions-heart, reactions-tada, interactions, created, updated |
commits |
author-date, committer-date |
code |
(unsupported) |
Instead of memorizing flags, save your frequently used or complex search conditions permanently as a .yaml file:
# One-command execution of a saved search profile
ghresearcher search --config examples/search_ai_repos.yamlYAML field names map one-to-one to the long CLI flags (underscore _ becomes hyphen -, e.g. good_first_issues ↔ --good-first-issues). The CLI-overrides-YAML rule: a CLI flag only overrides the corresponding YAML field when you explicitly pass it; otherwise the YAML value is used (defaults apply when there is no YAML).
Before actually calling gh search, GhResearcher runs two layers of validation to surface errors early:
- Field-name validation: an unknown field in the YAML (e.g. a typo like
langauge) prints a warning that the field may be ignored by gh. - Sort-value validation: an invalid
--sortvalue lists all valid values for that type.
Scenario A: Finding High-Quality Repositories (repos)
High-star Python repositories in a given field.
- YAML Config (
examples/search_ai_repos.yaml):item_type: repos query: "LLM agent" language: Python limit: 20 sort: stars order: desc topic: artificial-intelligence
- Equivalent CLI:
ghresearcher search repos "LLM agent" -L Python -t artificial-intelligence -s stars -O desc -l 20
Running output:
Searching repos for 'LLM agent'...
Running command: gh search repos LLM agent --limit 20 --sort stars --order desc --language Python --topic artificial-intelligence
melih-unsal/DemoGPT 🤖 Create LLM agents in a second with your prompts. Everything you need to create an LLM Agent - tools, prompts, frameworks, and models - all in one place. public 2026-08-24T14:42:12Z
zjunlp/MachineSoM [ACL 2024] Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology View public 2026-08-24T00:29:11Z
AkshitIreddy/AI-Plays-God-of-War LLM Agent paired with Image Captioning and Yolov8 models plays God of War public 2026-06-26T17:00:05Z
firelink-data/drive 🚀✨ DRIVE, the tool for creating and managing autonomous LLM agents; implemented using Apache Kafka, Docker, and LangChain. public 2025-03-07T17:54:37Z
jake12-cpu/AI-Desktop-Companion 基于 LLM Agent 的智能桌面陪伴系统，实现角色人格控制、长期记忆管理和自然语言交互。 public 2026-07-16T08:39:20Z
Atakan-Emre/McpTestGenerator Standardized Test Case Generation for LLM Agents via Model Context Protocol. Bridging the gap between AI and QA (Xray/Jira). public 2026-05-25T09:42:37Z
yuliu625/Yu-Agent-Development-Toolkit A robust, engineering-focused toolkit for building stable, complex LLM Agents. Built on the LangChain + LangGraph ecosystem for superior flow control and production readiness. public 2026-08-15T11:56:21Z
AbdulSamad502/InsightForge-AI Production-ready AI-powered Business Intelligence platform using LangGraph, LangChain, FastAPI, PostgreSQL, and LLM agents for automated data analysis, insights generation, forecasting, and reporting.ai-data-analyst syste, public 2026-08-15T08:21:57Z
mr-j90/PaperTrace An AI research assistant over the ~12,500 arXiv papers on RAG, LLM agents, LLM evaluation, and LLMOps — the literature about the very techniques it's built from. Ask it a question and watch it think: an agent visibly rewrites your query, chooses tools, gathers evidence across papers. public 2026-08-18T13:54:57Z
AnuragRoque/ExcelliaAI-MCP AI-Powered Spreadsheet Validation & Data Quality Platform Excellia AI is a local-first, AI-driven spreadsheet validation platform built to automate data cleaning, validation, anomaly detection, and enrichment for large datasets. It combines rule-based checks, machine learning, and local LLM agents to drastically reduce manual analyst effort. public 2026-08-15T10:10:44Z
varunbiluri/fastapi-tool-agent-clean Production LLM agent with FastAPI, Azure OpenAI, tool use, CI/CD & Key Vault public 2026-07-11T16:39:35Z
grimdalltech/MemLens Memory observability for LLM agents—trace what they remember, why, and how it shapes every response. public 2026-08-19T16:57:46Z
sunyifei-126/EvoGate-RSI Evidence-gated Recursive Self-Improvement (RSI) runtime for self-improving LLM agents, self-evolving agents, agentic AI, AI safety, evaluation, lineage, and rollback. public 2026-08-10T09:25:08Z
MarcoLombardoDev/Argus AI-powered desktop application for cryptocurrency forecasting, multi-agent market analysis, backtesting, portfolio management, and automated trading. Combines TimesFM 2.5, KNN pattern matching, LLM agents, quantitative analysis, and CCXT. public 2026-08-25T11:03:05Z
LPK3215/agentbase AI Agent 智能体脚手架 · LLM Agent 框架 / 项目脚手架. Configuration-driven AI Agent backend on deepagents + LangChain + LangGraph. YAML config, pluggable registries, CLI, FastAPI (21 routes), RAG knowledge base, MCP, queue, Docker. Build production-grade intelligent agents without boilerplate. public 2026-08-25T06:58:40Z
k0n1m4k1/kv-memory-modules Precompiled KV memory modules for LLM agents: compile Markdown memories once into relocatable, composable KV-state artifacts (.kmd) and link them at any position of a live context, no re-prefill. 7.0-27.6x faster session setup; validated on 9 models (2B-14B) over stock llama.cpp and vLLM. public 2026-08-25T20:21:47Z
ioanfesteu/multiagent_LLM Caretaker LLM agent meets Active Inference agents public 2026-02-23T17:35:25Z
virbahu/chatgpt-supply-chain-agent LLM agent framework for supply chain decision support and automation public 2026-06-08T02:33:53Z
VenkatLaxmi-code/eco-loop-building-agents Autonomous AI-powered smart building energy control system using EnergyPlus, LLM agents, MCP, and closed-loop optimization. public 2026-07-26T16:08:21Z
FarazIbrahim/agentic-due-diligence Multi-Agent AI Due Diligence Platform for Startup Investment Analysis using Multiple LLMs, Agentic Workflows, and Structured Evaluation Frameworks. public 2026-07-23T05:47:57Z
Scenario B: Exploring Codebase & Snippets (code)
Pinpoint legacy issues (e.g. TODO comments) inside a specific repo. Note: code search is powered by GitHub's legacy code search engine.
- YAML Config (
examples/search_code_todos.yaml):item_type: code query: "TODO" repo: "MaybeBio/GhResearcher" filename: "*.py" extension: "py" limit: 50
- Equivalent CLI:
ghresearcher search code "TODO" -r MaybeBio/GhResearcher -f "*.py" -e py -l 50
Scenario C: Tracking Issues & PRs (issues / prs)
Track recently merged bug fixes inside a large organization.
- YAML Config (
examples/search_bug_prs.yaml):item_type: prs query: "fix bug state:merged" owner: "microsoft" limit: 40 order: desc
- Equivalent CLI:
ghresearcher search prs "fix bug state:merged" -o microsoft -O desc -l 40
Scenario D: Structured JSON + jq Post-filtering (repos)
Get machine-readable fields and trim them with jq for scripts or pipelines.
# Output selected JSON fields
ghresearcher search repos "protein design" -L Python --json "fullName,stargazersCount,description" -l 10
# Keep only repo names via jq
ghresearcher search repos "protein design" -L Python --json "fullName,stargazersCount" --jq ".[] | .fullName" -l 10Scenario E: Presence Flags vs Value Booleans (prs / repos)
- Only merged draft PRs in Microsoft's org:
ghresearcher search prs "fix" -o microsoft --merged --draft -l 20 - Only non-archived public repositories (
--archivedis a value boolean, sofalsemust be explicit):ghresearcher search repos "deep learning" --archived false --visibility public -l 20
Scenario F: Pure Flag Filtering (omitting query)
repos/issues/prs/commits allow omitting keywords to list objects that only satisfy the filters.
ghresearcher search repos -o microsoft --visibility public -s stars -l 10For each item_type, we provide a fully-loaded template covering every option and JSON field. Keep only the parameters you need; delete or comment out the rest. These templates ship in the examples/ directory (template_*.yaml).
Two kinds of boolean flags (see §3.3):
- Value boolean (e.g.
archived): writearchived: falseorarchived: true, converted to--archived=false/true.- Presence flag (e.g.
draft/merged/no_assignee):trueappends--flag,falseomits it (never rendered as--flag=false).Note:
codesearch does not supportsortororder(gh search codehas neither flag natively), so they are omitted in the template; if written, they are ignored with a warning.
1. Repositories (repos)
item_type: repos
query: "machine learning" # Core keywords
# Result Controls
limit: 30
sort: stars # forks | help-wanted-issues | stars | updated
order: desc # asc | desc
# Boolean Flags
archived: false # value boolean, explicit true | false
include_forks: "false" # options: false | true | only
# Filter Qualifiers
language: python
topic: deep-learning
owner: MaybeBio
match: description
created: ">=2023-01-01"
followers: ">=5"
forks: ">=10"
good_first_issues: ">=5"
help_wanted_issues: ">=10"
license: "mit"
number_topics: ">=2"
size: "5000..10000"
stars: ">=100"
updated: ">=2023-01-01"
visibility: "public"
# Formatting Options
# web: true # open the search page in a browser
# json: ["id", "name", "owner", "stargazersCount", "description"]
# jq: ".[] | .name"
# template: "{{.name}}"2. Code (code)
item_type: code
query: "TODO" # code search requires a query
# Result Controls
limit: 30
# Note: code search does not support sort / order
# Filter Qualifiers
language: python
owner: MaybeBio
repo: MaybeBio/GhResearcher
extension: py
filename: main.py
match: file
size: "1..50"
# Formatting Options
# web: true # open the search page in a browser
# json: ["url", "path", "repository", "sha"]
# jq: ".[] | .path"
# template: "{{.path}}"3. Issues (issues)
item_type: issues
query: "crash label:bug is:open"
# Result Controls
limit: 30
sort: comments # comments | created | interactions | reactions | reactions-+1 | reactions--1 | reactions-heart | reactions-smile | reactions-tada | reactions-thinking_face | updated
order: desc # asc | desc
# Boolean Flags
archived: false # value boolean, explicit true | false
include_prs: false # presence flag
locked: false # presence flag
no_assignee: false # presence flag
no_label: false # presence flag
no_milestone: false # presence flag
no_project: false # presence flag
# Filter Qualifiers
app: "some-app"
assignee: "MaybeBio"
author: "MaybeBio"
closed: ">=2023-01-01"
commenter: "MaybeBio"
comments: ">=5"
created: ">=2023-01-01"
interactions: ">=10"
involves: "MaybeBio"
label: "bug"
language: python
match: title
mentions: "MaybeBio"
milestone: "v1.0"
owner: MaybeBio
project: "MaybeBio/1"
reactions: ">=5"
repo: MaybeBio/GhResearcher
state: "open"
team_mentions: "my-team"
updated: ">=2023-01-01"
visibility: "public"
# Formatting Options
# web: true # open the search page in a browser
# json: ["id", "title", "state", "url"]
# jq: ".[] | .title"
# template: "{{.title}}"4. Pull Requests (prs)
item_type: prs
query: "fix memory leak is:merged"
# Result Controls
limit: 30
sort: updated # comments | created | interactions | reactions | reactions-+1 | reactions--1 | reactions-heart | reactions-smile | reactions-tada | reactions-thinking_face | updated
order: desc # asc | desc
# Boolean Flags
archived: false # value boolean, explicit true | false
draft: false # presence flag
locked: false # presence flag
merged: true # presence flag
no_assignee: false # presence flag
no_label: false # presence flag
no_milestone: false # presence flag
no_project: false # presence flag
# Filter Qualifiers
app: "some-app"
assignee: "MaybeBio"
author: "MaybeBio"
base: "main"
checks: "success" # pending | success | failure
closed: ">=2023-01-01"
commenter: "MaybeBio"
comments: ">=5"
created: ">=2023-01-01"
head: "feature-branch"
interactions: ">=10"
involves: "MaybeBio"
label: "bug"
language: cpp
match: body
mentions: "MaybeBio"
merged_at: ">=2023-01-01"
milestone: "v1.0"
owner: microsoft
project: "microsoft/1"
reactions: ">=5"
repo: microsoft/terminal
review: "approved" # none | required | approved | changes_requested
review_requested: "MaybeBio"
reviewed_by: "MaybeBio"
state: "closed"
team_mentions: "my-team"
updated: ">=2023-01-01"
visibility: "public"
# Formatting Options
# web: true # open the search page in a browser
# json: ["id", "title", "state", "url"]
# jq: ".[] | .title"
# template: "{{.title}}"5. Commits (commits)
item_type: commits
query: "Initial commit"
# Result Controls
limit: 30
sort: author-date # author-date | committer-date
order: desc # asc | desc
# Boolean Flags
merge: false # presence flag
# Filter Qualifiers
author: "MaybeBio"
author_date: ">=2023-01-01"
author_email: "test@example.com"
author_name: "John Doe"
committer: "MaybeBio"
committer_date: ">=2023-01-01"
committer_email: "test@example.com"
committer_name: "John Doe"
hash: "8dd03144ffdc6c0d486d6b705f9c7fba871ee7c3"
owner: MaybeBio
parent: "parent_hash"
repo: MaybeBio/GhResearcher
tree: "tree_hash"
visibility: "public"
# Formatting Options
# web: true # open the search page in a browser
# json: ["id", "commit", "author", "url"]
# jq: ".[] | .commit.message"
# template: "{{.commit.message}}"- Exact Phrases: Use double quotes for exact phrase matching (e.g.
"memory leak"). - Boolean Logic: Qualifiers support
OR, and exclusion via a leading hyphen-(e.g.bug OR error,-wip). Warning: If your query itself starts with-, the shell may misinterpret it as a flag (Unix: use--before the query; PowerShell: use--%). - Domain Differences: Available filters are not universal across
item_type. For example,--topiconly works onrepos,--extensiononly oncode, and--sort/--orderare unsupported forcode. - Two Kinds of Boolean Flags: Value booleans (e.g.
archived) must be written explicitly astrue/false; presence flags (e.g.draft/merged/no_assignee) append--flagwhentrueand are omitted whenfalse— never rendered as--flag=false. - Built-in Validation: The tool validates field names and
sortvalues peritem_type. Unknown fields or invalidsortvalues print a yellowWarningbut do not abort execution (parameters are still passed through togh). - CLI overrides YAML: When the same parameter appears in both, the CLI value overrides the YAML value (other YAML parameters still apply).
- When
queryis required:repos/issues/prs/commitsallow omittingquery(pure flag filtering), butcodesearch requires aquery. - References: For the full qualifier-to-flag mapping tables, see the official GitHub manuals collected under
docs/(docs_github_com_en_search-github_*).
Here we use the monitor command as an example to show how to push GhResearcher's output to you automatically:
As mentioned earlier, I have some target users and organizations whose latest activity I want to see every day, especially the content they follow (files and commands below). It would be great if this could be fetched automatically every day — like scrolling a social feed — so I don't have to manually run the query each time.
# you can change the input file to your own target file
# yesterday-to-today events of followed users
ghresearcher monitor -f /data2/GhResearcher/tests/target_academic_user.txt --since $(date -d "1 day ago" +%Y-%m-%d) --expand-commits
# yesterday-to-today events of followed users (including received events; this output is usually very long)
ghresearcher monitor -f /data2/GhResearcher/tests/target_academic_user.txt --since $(date -d "1 day ago" +%Y-%m-%d) -r --expand-commits
# yesterday-to-today events of followed organizations
ghresearcher monitor -f /data2/GhResearcher/tests/target_org.txt --since $(date -d "1 day ago" +%Y-%m-%d) --org --expand-commitsSince I frequently push updates to the user/organization lists above (like expanding my social circle), here is a single simple command to run locally:
curl -f -o protein_dl_user.txt https://raw.githubusercontent.com/MaybeBio/GhResearcher/refs/heads/main/tests/protein_dl_user.txt && ghresearcher monitor -f protein_dl_user.txt --since $(date -d "1 day ago" +%Y-%m-%d) --expand-commits && rm -f protein_dl_user.txt
⚠️ ghresearcher monitor -fcurrently only accepts a file path on disk, not piped input
- Alias (fixed command): write the command above into
~/.bashrcor~/.zshrcas a fixed alias for daily use. However, it can't handle positional parameters — if you ever want to look back two days, you'd have to manually edit the--sinceparameter in the command. - Shell function: wrap it into a shell function that accepts parameters and write it into
~/.bashrcor~/.zshrc. You can also save the output as a log for reviewing hot activity later. Example:
ghfollow() {
# look back 1 day by default
local days=1
# pass an argument like `ghfollow -2` to look back 2 days
# any argument overrides the default and is parsed into N days
if [[ $# -ge 1 ]]; then
days="${1#-}"
fi
# a unique temp file under /tmp, removed afterwards
local tmp_raw=$(mktemp)
local log_dir="$HOME/ghfollow_log"
# archive logs by date: when it was recorded and what period it covers
mkdir -p "${log_dir}"
local date_folder=$(date +%Y%m%d)
local target_log_dir="${log_dir}/${date_folder}"
mkdir -p "${target_log_dir}"
# down to hour/minute/second so you can run multiple times a day
# (though gh output already has timestamps, so time isn't strictly needed)
# $HOME/ghfollow_log/2026-08-13/090000_past1days.log
local logfile="${target_log_dir}/$(date +%H%M%S)_past${days}days.log"
# run the command and save the log
curl -f -o "${tmp_raw}" https://raw.githubusercontent.com/MaybeBio/GhResearcher/refs/heads/main/tests/protein_dl_user.txt && \
ghresearcher monitor -f "${tmp_raw}" --since "$(date -d "${days} day ago" +%Y-%m-%d)" --expand-commits 2>&1 | tee "${logfile}"
rm -f "${tmp_raw}"
echo "✅ log saved: ${logfile}"
}The resulting directory structure looks like:
~/ghfollow_log/
├─20260813/
│ ├─090000_past1days.log
│ └─142231_past3days.log
└─20260814/
└─090000_past1days.log
- crontab / systemd scheduled tasks: schedule the shell function above via crontab or a systemd timer to fetch activity and save logs automatically.
- tmux / screen persistent sessions (print to a resident terminal): the previous approach (crontab/systemd) can't output directly to an interactive shell — they are background daemons with their own sessions, so output can only be written to a file and reviewed later, not printed live to your current SSH terminal. If you really want output printed in the shell at, say, 9:00 AM, use
tmuxorscreenfrom within the scheduled task: run atmux send-keyscommand that sends the shell function to a designatedtmuxsession, so that whenever you attach to that tmux window you see the output in the current terminal. - Scheduled execution + persistent Git storage: the best approach is GitHub Actions — write a workflow that runs the command above on a schedule and commits the output back to the repository, giving you both scheduled execution and persistent storage.
In short, my personal approach is: for simple local use, no scheduled recording is needed (just wrap a function in
bashrc/zshrc); for long-term scheduled tasks, GitHub Actions is the key (recommended for persistent records).
- The
monitor/parsecommands are basically solid, butsearchstill needs more testing and optimization — especially YAML config parsing and the CLI-overrides-YAML logic, ensuring correct behavior across all combinations, and making it as concise and user-friendly as possible compared togh search. - The README has only been updated in Chinese for now; the English version will be synced in a later update.
- Further extend
parse: for repository parsing, consider integrating or borrowing from existing code-analysis tools to provide richer results, such as dependency graphs, call graphs, AST analysis, etc. - Fix the Rich output column-width issue — when output is too long, Rich tables may be truncated or wrapped, hurting readability; prefer one full line per row without wrapping.
- Optimize further by publishing/implementing as a
GitHub CLI extension, so users can install viagh extension installwithout a separate Python environment. See: https://github.com/topics/gh-extension | https://cli.github.com/manual/
- GitHub API Limits: The
monitortimeline is restricted to a maximum of 300 recent events or events within the past 90 days due to GitHub API constraints. - Parser Size: The
parsecommand can still be expensive for extremely large repositories because tree rendering and pager output may be large. When the Trees API is truncated, the tool falls back to a temporary shallow clone. - Rate Limiting: Heavy use of
--expand-commitsor batch mapping large lists may quickly consume your GitHub API rate limit. Use with care.