r3bl-open-core / task
AGENTS.md486 starsChanged 15 days agoApache-2.0View on GitHub
Tasks are instructions to guide claude code to implement things. One task per .md file. Keep it concise: - Use [x]/[ ] checkboxes, not status codes on headings. - Use ### phase headers to group related steps. Only add phases when steps naturally cluster. - Mandatory Manual Review: Every task, phase, and sub-phase MUST end with a "Mandatory manual review" step containing a checkbox for each modified file. CRITICAL: These checkboxes MUST be left as [ ] by the agent.…
We split this AGENTS.md into the rules it gives an agent
6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
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 34
Use [x]/[ ] checkboxes, not status codes on headings.
- Very specific: line 35
Use ### phase headers to group related steps.
- Distinctive: line 36
Mandatory Manual Review: Every task, phase, and sub-phase MUST end with a "Mandatory manual review" step containing a checkbox for each …
- Distinctive: line 40
Prefer one line per step. A brief technical note or sub-bullet is fine for clarity.
- Very specific: line 41
Do NOT use doctoc, numbered step headings (Step 0.0), or per-step detail blocks.
- Very specific: line 44
Formatting: Always run prettier --write <file> on any markdown files after creating or editing them to ensure they are properly formatted …
≈ “Common” is an estimate: at least 1 in 5 of the 36,368 AGENTS.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
Rules for managing tasks
Tasks are instructions to guide claude code to implement things. One task per .md file.
Structure of a task file
# Task: [Short Name]
## Overview
Why this task exists and what it accomplishes. Include architectural context only if
non-obvious from the code.
## Implementation plan
### Phase 1: [Phase name]
- [x] Completed step
- [ ] Pending step
- [ ] Another pending step
### Phase 2: [Phase name]
- [ ] Step with brief technical note if needed
- [ ] Another step
- [ ] **Mandatory manual review:** Verify every file modified in this phase for correct
implementation and ensure no regressions.
- [ ] `path/to/modified_file_1.rs`
- [ ] `path/to/modified_file_2.rs`Keep it concise:
- Use
[x]/[ ]checkboxes, not status codes on headings. - Use
###phase headers to group related steps. Only add phases when steps naturally cluster. - Mandatory Manual Review: Every task, phase, and sub-phase MUST end with a
"Mandatory manual review" step containing a checkbox for each modified file.
CRITICAL: These checkboxes MUST be left as
[ ]by the agent. Only the user may check them after their manual review is complete. - Prefer one line per step. A brief technical note or sub-bullet is fine for clarity.
- Do NOT use doctoc, numbered step headings (
Step 0.0), or per-step detail blocks. - Exact Signatures & File Paths for Implementation: Include exact Rust type signatures, trait bounds, and target file paths in the plan so implementation steps can be executed deterministically without guessing architecture. Avoid large boilerplate implementation blocks.
- Technical notes and context belong in the Overview, not repeated per step.
- Formatting: Always run
prettier --write <file>on any markdown files after creating or editing them to ensure they are properly formatted and easy to review in the user's IDE.
Folder structure
task/- Active tasks (currently being worked on).task/pending/- Future tasks (not yet started).task/done/- Completed tasks.task/archive/- Abandoned tasks kept for historical reference.
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 28,693 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository30 other files it gives its agents.
- .agents/AGENTS.mdAGENTS.md
- AGENTS.mdAGENTS.md
- build-infra/AGENTS.mdAGENTS.md
- rust-analyzer-mcp-server/AGENTS.mdAGENTS.md
- llms.txtllms.txt
- analyze-log-filesSkill
- analyze-performanceSkill
- audit-taskSkill
- batch-refactor-with-sub-agentsSkill
- check-bounds-safetySkill
- check-code-qualitySkill
- check-test-coverageSkill
- code-reviewSkill
- concurrency-safetySkill
- create-commit-messageSkill
- create-prSkill
- design-philosophySkill
- fast-string-allocationsSkill
- fix-issueSkill
- merge-prSkill
- organize-modulesSkill
- organize-testsSkill
- release-crateSkill
- remove-crate-prefixSkill
- review-prSkill
- run-clippySkill
- safe-renamingSkill
- test-cross-platformSkill
- write-documentationSkill
- write-structured-tracingSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.

