Engineering-Bible-AI
AGENTS.md0 starsChanged 4 months agoMITView on GitHub
This repository packages portable engineering instructions, standards, skills, templates, validation, and installers. Treat it as a public distribution: local runtime state and private infrastructure do not belong in tracked files. Select the narrowest directly relevant skill. Do not invoke workflow-router for a clear single-domain request; reserve it for ambiguous or genuinely mixed work. Use one primary skill and at most one supporting skill by default. For a same-task follow-up in the same thread, reuse the current skill route and already-loaded…
We split this AGENTS.md into the rules it gives an agent
6 rules. About 1 is common, 4 are distinctive, and 1 is very specific to this project.
≈1
Common · estimate
Many files like this have a rule of this kind.
≈4
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 39
Do not commit credentials, auth files, provider endpoints, environment files, private keys, local agent sessions, caches, generated …
- Distinctive: line 55
Prefer the smallest correct change.
- Very specific: line 60
Use apply_patch for hand-authored edits.
- Distinctive: line 62
Never claim a check passed unless the exact command ran successfully.
- Distinctive: line 66
Use the smallest profile that proves the change, then run the broader profile when shared tooling, installation, or release behavior …
- Distinctive: line 82
Before completion, review the diff for public behavior, runtime-boundary drift, generated documentation drift, missing regression coverage, …
≈ “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
Engineering Bible AI Repository Instructions
Purpose
This repository packages portable engineering instructions, standards, skills, templates, validation, and installers. Treat it as a public distribution: local runtime state and private infrastructure do not belong in tracked files.
Routing Discipline
Select the narrowest directly relevant skill. Do not invoke workflow-router
for a clear single-domain request; reserve it for ambiguous or genuinely mixed
work. Use one primary skill and at most one supporting skill by default.
Continuation Fast Path
For a same-task follow-up in the same thread, reuse the current skill route and
already-loaded instructions. Do not route again or reread an unchanged
SKILL.md unless the domain, risk, required tools, or requested workflow
changed. If compaction removed required instructions, reload only the missing
skill.
Repository Sources Of Truth
skills/registry.ymlowns skill membership, grouping, and order.- Generated registry blocks in
README.md,README.ru.md, andMANIFEST.mdmust be updated withpython3 scripts/registry.py --root . docs --write; do not edit their contents by hand. engineering/README.mdindexes the standards library. Add or remove a standard and its index entry together.instructions/global/steady.mdis the default installed global prompt.instructions/global/full.mdpreserves exhaustive routing for strict use;instructions/global/minimal.mdis the compact steady-state profile.- Root
AGENTS.mdcontains project-only rules and must not duplicate either global prompt.
Runtime Boundary
Do not commit credentials, auth files, provider endpoints, environment files, private keys, local agent sessions, caches, generated capability catalogs, or machine-specific configuration. Runtime tools are discovered dynamically from the current session; tracked policies must describe capabilities and risk, not hard-code provider or tool identifiers.
Generated project-local capability files belong under .engineering-bible/
and must remain untracked. Do not claim a runtime capability is available until
the current session exposes it.
Change Rules
- Inspect relevant implementation, tests, docs, and current Git state before editing.
- Preserve unrelated dirty-worktree changes; never reset, stash, or overwrite them without explicit authorization.
- Prefer the smallest correct change. Keep installer, validation, registry, runtime discovery, and documentation responsibilities separate.
- Public CLI changes require compatibility handling, tests, and matching docs.
- Validation output must distinguish
PASS,FAIL, andSKIP. A skipped release gate is not success. - Use
apply_patchfor hand-authored edits. Generated-block updates and formatting may use their deterministic project commands. - Never claim a check passed unless the exact command ran successfully.
Validation
Use the smallest profile that proves the change, then run the broader profile when shared tooling, installation, or release behavior changes:
make validate-quick
make validate-bootstrap
make validate
make validate-releaseTests are discovered with:
python3 -m unittest discover -s tests -p 'test_*.py' -vBefore completion, review the diff for public behavior, runtime-boundary drift, generated documentation drift, missing regression coverage, and accidental changes outside scope. Report exact commands, outcomes, assumptions, and remaining risks.
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 repository62 other files it gives its agents.
- accessibility-ui-reviewSkill
- agent-retrospectiveSkill
- agents-md-retrospectiveSkill
- agent-squadSkill
- architecture-mapSkill
- architecture-normalizerSkill
- architecture-principlesSkill
- authz-boundary-reviewSkill
- c-cppSkill
- code-qualitySkill
- code-reviewSkill
- code-to-figmaSkill
- code-wiki-ruSkill
- context-packSkill
- core-engineeringSkill
- debuggingSkill
- dependency-advisory-auditSkill
- deserialization-parser-reviewSkill
- design-system-extractorSkill
- documentationSkill
- engineering-standardsSkill
- esp32Skill
- esphomeSkill
- external-agent-pack-auditSkill
- fastSkill
- figma-to-codeSkill
- fix-security-findingSkill
- goSkill
- homeassistantSkill
- mcp-tool-routerSkill
- migration-plannerSkill
- mobile-qaSkill
- multi-agent-pr-reviewSkill
- performanceSkill
- playwright-visual-qaSkill
- pythonSkill
- quality-gatesSkill
- refactoringSkill
- responsive-breakpoint-checkSkill
- review-routerSkill
- rustSkill
- secrets-and-config-reviewSkill
- security-diff-reviewSkill
- security-routerSkill
- securitySkill
- session-memorySkill
- specialist-dispatchSkill
- subagent-result-mergeSkill
- supply-chain-reviewSkill
- tdd-guardSkill
- testing-tddSkill
- threat-modelSkill
- typescriptSkill
- ui-buildSkill
- ui-business-appsSkill
- ui-concept-firstSkill
- ui-figmaSkill
- ui-qaSkill
- ui-researchSkill
- ui-routerSkill
- The first 60 shown.
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.

