literary-critic
CLAUDE.md0 starsChanged 27 days agoMITView on GitHub
Were you spawned by another session, with a task already assigned? Then skip straight to that task. Steps 2-5 below need the GitHub API, and a spawned session has no mcp__* tools — implement, run the gate, and let claude-arsenal/bin/open_task_pr.sh push. It prints branch:<name> when it cannot open the PR here; return that line and stop. Never claim or release. Otherwise, every session, without waiting to be asked: 1. Read arsenal/session/handover.md for the previous session's context. 2. List the repository's…
In this repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- claude-arsenal/AGENTS.mdAGENTS.md
- coverage-gapsSkill
- dep-upgradeSkill
- designSkill
- executionSkill
- explore-ideaSkill
- gate-checkSkill
- githubSkill
- initSkill
- litcriticSkill
- mutmut-reportSkill
- pin-checkSkill
- pypi-releaseSkill
- python-bootstrapSkill
- queue-addSkill
- queue-nextSkill
- queue-statusSkill
- reviewSkill
- session-endSkill
- shipSkill
- specifySkill
The file
We split this CLAUDE.md into the rules it gives an agent
4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
Read arsenal/session/handover.md for the previous session's context.
- Very specific: line 18
Run python3 claude-arsenal/scripts/query_status.py --issues <that file> for the board, and report anything it flags.
- Very specific: line 25
Every change goes through a PR, ad hoc requests included — never push to the default branch.
- Very specific: line 30
After any session with tasks: update arsenal/session/handover.md.
≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- claude-arsenal: auto-managed -->Automatic session protocol
Were you spawned by another session, with a task already assigned? Then skip
straight to that task. Steps 2-5 below need the GitHub API, and a spawned session
has no mcp__* tools — implement, run the gate, and let
claude-arsenal/bin/open_task_pr.sh push. It prints branch:<name> when it
cannot open the PR here; return that line and stop. Never claim or release.
Otherwise, every session, without waiting to be asked:
- Read
arsenal/session/handover.mdfor the previous session's context. - List the repository's issues labelled
arsenal:task— open and closed — and save the JSON. Use whatever GitHub access this surface has; runclaude-arsenal/bin/github_channel.sh --detectto find out which. Requestnumber,title,state,labels,assigneesand notbody— the bodies are the bulk of that fetch and nothing downstream reads them. - Run
python3 claude-arsenal/scripts/query_status.py --issues <that file>for the board, and report anything it flags. - Pick up work:
python3 claude-arsenal/scripts/task_select.py --issues <that file>returns the next unblocked task, thenbash claude-arsenal/bin/claim_task.sh <id>takes it (see@claude-arsenal/AGENTS.md).- Nothing returned + workspace plans exist → seed tasks from each plan.
- Nothing at all → ask what to work on.
- Every change goes through a PR, ad hoc requests included — never push to the default
branch. A task's PR carries
Closes #<issue>so merging it closes the task by itself. Dispatching the work to another session instead? Pass the repository explicitly and passARSENAL_TASK_ISSUE— a spawned worker can resolve neither. →claude-arsenal/references/orchestrator-tick.md - After any session with tasks: update
arsenal/session/handover.md.
Specs and plans live at status/specification.md and status/plan.md (a workspace's at
arsenal/project/<ws>/spec.md / plan.md). A brainstorming or planning skill from another
plugin writes its output there too, not to its own default location.
@claude-arsenal/AGENTS.md
<!-- /claude-arsenal: auto-managed -->Project rules
@AGENTS.md
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

