sdlc-skills
AGENTS.md2 starsChanged 5 days agoMITView on GitHub
Welcome! You are working on the sdlc-skills repository, which houses the universal Spec-Driven, DDD & TDD Software Development Life Cycle agent skills. This repository uses Changesets for Semantic Versioning and automated changelog generation:
- Commits and pushes
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 31,774 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.
- adr-amendSkill
- adr-createSkill
- adr-review-alternativesSkill
- adr-validateSkill
- agents-maintainerSkill
- audit-docsSkill
- explore-domainSkill
- sdlc-guideSkill
- setup-sdlcSkill
- inception-workshopSkill
- prd-workshopSkill
- user-story-mappingSkill
- modify-flowSkill
- reverse-engineer-domainSkill
- create-moduleSkill
- implement-flowSkill
- create-entity-lifecycleSkill
- create-featuresSkill
- create-flow-documentationSkill
The file
We split this AGENTS.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 21
Every workflow/template-producing skill MUST open with XML control-plane blocks (<state_machine>, <execution_invariants>, …
- Very specific: line 36
Documenting changes: Run npm run changeset when adding or modifying skills, choose the bump level (patch, minor, major), and write a …
- Very specific: line 37
Bumping version & changelog: Run npm run version. This will:
- Very specific: line 55
Local setups: Run bash scripts/link-skills.sh after a git pull to refresh symlinks in ~/.claude/skills and ~/.agents/skills.
≈ “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
SDLC Skills Guidelines for Agents
Welcome! You are working on the sdlc-skills repository, which houses the universal Spec-Driven, DDD & TDD Software Development Life Cycle agent skills.
Architecture & Conventions
- Each skill lives in
skills/<category>/<skill-name>/with aSKILL.mdentry point. SKILL.mdMUST contain standard YAML frontmatter:--- name: <skill-name> description: >- <Short description and trigger phrases> disable-model-invocation: true # or false for router skills ---- Subdirectories inside a skill can include
templates/,references/, andscripts/. - Shared guidelines reside in
guidelines/and are symlinked into categories. - Skills must remain modular, decoupled from any single proprietary codebase, and configurable via
.sdlc/config.json. - Level 3 Execution Protocol (
guidelines/execution-protocol.md):- Every workflow/template-producing skill MUST open with XML control-plane blocks (
<state_machine>,<execution_invariants>,<output_contract>) that enforceworkflow structure = immutable | step reasoning = flexible. - Every reference to a file in
templates/*.mdMUST be locked inside<output_contract>requiring an explicitread_filetool call on the template before generating output (never reconstructing templates from prose or memory), plus an explicitESCAPE_HATCH("N/A — [1-line rationale]"or ask the user; never silently omit or invent sections). - Multi-stage skills MUST enforce explicit
🛑 HALT TURNboundaries (with an explicit fast-track override when the user requests batch/Light execution) and end with an evidence-based<self_validation_gate>block that requires citing exact tool calls, file paths, and section counts rather than passive boolean checkboxes.
- Every workflow/template-producing skill MUST open with XML control-plane blocks (
Verifying Skills
Run:
bash scripts/list-skills.shVersioning, Changelog & Releases
This repository uses Changesets for Semantic Versioning and automated changelog generation:
- Documenting changes: Run
npm run changesetwhen adding or modifying skills, choose the bump level (patch,minor,major), and write a summary. - Bumping version & changelog: Run
npm run version. This will:- Consume pending changesets.
- Update
package.jsonversion. - Update
CHANGELOG.md. - Automatically sync the version into
.claude-plugin/plugin.jsonand marketplace manifests (.claude-plugin/marketplace.json,.agents/plugins/marketplace.json) viascripts/sync-versions.js.
- Commit, Tag & Publish:
git add . git commit -m "chore(release): vX.Y.Z" git tag vX.Y.Z git push origin main --tags
How Different Agents Consume Updates
- Claude Code: Reads
.claude-plugin/plugin.json(auto-synced) when users runclaude plugin update sdlc-skills. - GitHub Copilot & Codex: Track the repository and git tags directly via their respective plugin marketplace settings.
- Universal / Antigravity / Cursor / Windsurf: Installers like
npx skills@latest add faparicior/sdlc-skillspull the latest git release/tag. - Local setups: Run
bash scripts/link-skills.shafter agit pullto refresh symlinks in~/.claude/skillsand~/.agents/skills.
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.

