agentic-developer-platform
AGENTS.md7 starsChanged 6 days agoApache-2.0View on GitHub
This is an open-source repository. Follow docs/PUBLISHING.md: never publish real AWS account IDs, deployment URLs, customer/environment identities, resource IDs, operator profiles, or private evidence in documentation. Use fictional examples and private configuration. Run python3 scripts/check-public-docs.py before publishing documentation changes. Follow the canonical issue-authoring guide and its epic/story templates. Preserve native parent links and stable acceptance and scenario IDs; specify independent test assertions and relevant determinism controls. The Engineering Insights walkthrough provides filled examples and an optional offline format check. To…
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 31,843 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 repository20 other files it gives its agents.
- modules/agent-factory/AGENTS.mdAGENTS.md
- modules/domain-apps/superplane/AGENTS.mdAGENTS.md
- modules/L5-services/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- evidence-skillSkill
- adp-flowSkill
- grillingSkill
- aidlc-emit-issuesSkill
- codex-bridgeSkill
- stage-1-triageSkill
- stage-2-osintSkill
- stage-3-staticSkill
- stage-4-dynamicSkill
- stage-5-correlationSkill
- stage-6-verdictSkill
- stage-7-reportSkill
- url-analysisSkill
- url-analysisSkill
- skypilotSkill
- superplaneSkill
The file
We split this AGENTS.md into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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
- Very specific: line 8
Use fictional examples and private configuration.
- Distinctive: line 13
Follow the canonical issue-authoring guide and its epic/story templates.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 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
Agent Instructions — ADP (Agentic Developer Platform)
Public documentation
This is an open-source repository. Follow docs/PUBLISHING.md:
never publish real AWS account IDs, deployment URLs, customer/environment
identities, resource IDs, operator profiles, or private evidence in documentation.
Use fictional examples and private configuration. Run
python3 scripts/check-public-docs.py before publishing documentation changes.
Epic and story authoring
Follow the canonical issue-authoring guide and its epic/story templates. Preserve native parent links and stable acceptance and scenario IDs; specify independent test assertions and relevant determinism controls. The Engineering Insights walkthrough provides filled examples and an optional offline format check.
Deployment
To deploy ADP, read and follow the canonical agent-deploy guide:
docs/adp-platform-deployment/deploy-with-agent.md
It is the single source of truth for the agent-behavior layer (confirm the
target account, the phase table, the "placeholder artifact" rule, state file,
verification, teardown, when to call the user). It in turn defers to
docs/adp-platform-deployment/deploy-quickstart.md
for the exact, verified commands.
The essentials, so you don't start down the wrong path:
- Confirm the target AWS account first (
aws sts get-caller-identityvia the activeAWS_PROFILE) and get the user's OK before Phase 1. - There is no upfront GitHub setup. GitHub is wired at the END for the agent
path (UI flow: Settings → Connections → "Set up GitHub App"; or CLI fallback
register-github-app.sh). Any "Phase 0 / setup-org / 3 org-owned apps" instruction is the superseded legacy ARC track — do not run it. - The full launcher chains core deployment stages, including broker, first-admin bootstrap, webhook, separate agent factory and final frontend publication. Follow the canonical guide for ordering and scope. GitHub App browser wiring comes after deployment.
The deployment instructions are intentionally a redirect so the procedure has one source of truth and never drifts across the agent-entry files.
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.