suwappubot
AGENTS.md4 starsChanged 3 months agoApache-2.0View on GitHub
Policy for AI agents (Claude Code and others) working in this repository. Precedence: this file → ARCHITECTURE.md → CONVENTIONS.md. CLAUDE.md is the Claude-Code-specific operational profile (conductor protocol, routing table, response-length rules) and is kept authoritative for those details — this file is the tool-agnostic summary. 1. Implement, don't plan. Bounded exploration, then build. Plans are deliverables only when explicitly requested. 2. Evidence over assertion. Done-ness claims cite command exits or live-URL checks. bash scripts/verify.sh aggregates the gates; python3 scripts/status.py checks…
- Reads credentials
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 33,315 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 repository19 other files it gives its agents.
- .codex/AGENTS.mdAGENTS.md
- api-ts/CLAUDE.mdCLAUDE.md
- bot/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- showcase/CLAUDE.mdCLAUDE.md
- webapp/CLAUDE.mdCLAUDE.md
- gitbook/llms.txtllms.txt
- suwappubot-conventionsSkill
- aegis-goalSkill
- design-iterateSkill
- ethglobal-goalSkill
- evolveSkill
- goalSkill
- handoffSkill
- redesign-existing-projectsSkill
- reflectSkill
- suwappubot-conventionsSkill
- design-taste-frontendSkill
- suwappuSkill
The file
We split this AGENTS.md into the rules it gives an agent
2 rules. About 1 is common, 0 are distinctive, and 1 is very specific to this project.
≈1
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Common: line 22
Secrets: never print values (SET/UNSET only), never commit env files, never widen a secret's scope to make a test pass.
- Very specific: line 46
Read before re-deriving: docs/README.md (index of all docs, staleness-flagged), docs/architecture/OVERVIEW.md (services, ~23 background …
≈ “Common” is an estimate: at least 1 in 5 of the 42,283 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
AGENTS.md — executable agent policy
Policy for AI agents (Claude Code and others) working in this repository. Precedence:
this file → ARCHITECTURE.md → CONVENTIONS.md. CLAUDE.md is the Claude-Code-specific
operational profile (conductor protocol, routing table, response-length rules) and is
kept authoritative for those details — this file is the tool-agnostic summary.
Ground rules
- Implement, don't plan. Bounded exploration, then build. Plans are deliverables only when explicitly requested.
- Evidence over assertion. Done-ness claims cite command exits or live-URL checks.
bash scripts/verify.shaggregates the gates;python3 scripts/status.pychecks what's actually running;python3 scripts/doctor.pyprobes the local toolchain and capability state without printing secrets. - Money-path changes get adversarial review. Anything touching swap execution,
wallet/keys, KMS, billing/x402, fee math, points, or withdrawals is tagged
MONEY-PATHand reviewed by the dedicated reviewer before merge. - Never rebase; never add Co-Authored-By; black before push. See
CONVENTIONS.md. - Generated artifacts are regenerated, not edited:
.env.schema,api-ts/openapi-agent.json. - Secrets: never print values (SET/UNSET only), never commit env files, never widen a secret's scope to make a test pass.
Repo skills (reuse before building)
Workflows exist as skills (.claude/commands/) — use them instead of hand-rolling:
/ship (branch→PR→CI→merge→boot verify), /deploy, /status, /audit,
/audit-fleet, /bugclass, /worktree-check. For scaffolding (new handler,
route, page) copy an existing sibling as the template; dual-ORM schema changes
follow docs/development/migrations.md.
Where things live
| Task | Entry point |
| Telegram command | bot/handlers/ + registration; logic in bot/services/ |
| api-ts endpoint | api-ts/src/routes/ (Hono + Effect-TS) |
| Schema change | Both bot/models/ (SQLAlchemy) and api-ts Drizzle; additive+idempotent |
| Webapp page | webapp/src/pages/ |
| Env var | bot/config/settings.py or api-ts/src/config/EnvService.ts, then regenerate .env.schema |
| Optional provider | Declare in capabilities.yaml; the manifest, not the code, is the truth source |
Knowledge base
Read before re-deriving: docs/README.md (index of all docs, staleness-flagged),
docs/architecture/OVERVIEW.md (services, ~23 background tasks, request flows),
docs/ONBOARDING.md (setup/env/test facts), docs/DECISIONS.md (lessons log),
docs/adr/ (append-only decision records — link one on MONEY-PATH/cross-stack PRs).
Component-scoped rules: bot/CLAUDE.md, api-ts/CLAUDE.md, webapp/CLAUDE.md
(read the one for the directory you're editing).
Verification ladder (cheapest first)
- Parse gate:
python3 -c "import ast; ast.parse(open('file.py').read())" - Type gate:
bun run check(api-ts, incremental) - Unit gates:
pytest tests/, webappnpm run test - Aggregate:
bash scripts/verify.sh - Boot gate:
python3 scripts/status.py+ Railway logs clean of import errors - Live gate: real end-to-end exercise of the changed path
CI green passes gates 1–4 only. A deploy claim requires 5; a "feature works" claim requires 6.
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.

