domain-build
Skill8 starsChanged 6 months agoMITView on GitHub
Use when building, deploying, and documenting skills from a completed domain-design output. Handles the walking skeleton -> remaining P0s -> cookbook -> handoff cycle.
In this repository129 other files it gives its agents.
- AGENTS.mdAGENTS.md
- guidance/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- guidance/CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- blast-radius-checkSkill
- brainstormingSkill
- branch-flow-gateSkill
- branch-sync-gateSkill
- codebase-reconSkill
- codeowners-drift-auditSkill
- code-review-batterySkill
- cognitive-complexity-refactoringSkill
- debateSkill
- debug-conductorSkill
- dispatching-parallel-agentsSkill
- evidence-adjudicatorSkill
- executing-plansSkill
- external-cli-auditSkill
- feature-developmentSkill
- field-rename-verificationSkill
- finishing-a-development-branchSkill
- git-branch-conventionsSkill
- gitlab-cliSkill
- hotfix-charterSkill
- implementation-trackerSkill
- incident-revert-reviewSkill
- infra-config-investigatorSkill
- investigation-stateSkill
- kernel-splitSkill
- llm-behavior-investigatorSkill
- llm-skill-reviewSkill
- merge-authorization-gateSkill
- micro-harsh-reviewSkill
- output-verificationSkill
- pre-commit-gateSkill
- progressive-code-review-gateSkill
- progressive-harsh-reviewSkill
- providing-code-reviewSkill
- pr-triage-gateSkill
- push-authorization-gateSkill
- receiving-code-reviewSkill
- reproduction-experiment-investigatorSkill
- requesting-code-reviewSkill
- requirements-validation-pmSkill
- requirements-validationSkill
- scope-tripwireSkill
- session-handoffSkill
- skills-hierarchy-tuningSkill
- sp-bughuntSkill
- state-consistency-investigatorSkill
- subagent-driven-developmentSkill
- systematic-debuggingSkill
- test-driven-developmentSkill
- timeline-trace-investigatorSkill
- token-estimationSkill
- unified-commit-gateSkill
- using-git-worktreesSkill
- using-superpowersSkill
- verification-before-completionSkill
- The first 60 shown.
The file
Show its rules and text
We split this skill into the rules it gives an agent
13 rules. About 0 are common, 10 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈10
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 34
Use todo-management skill for all 4 phases.
- Very specific: line 46
Create skills/{domain}/{skill-name}/skill.md with frontmatter:
- Distinctive: line 58
Verify the skill is discoverable: run the discovery command for your skill system.
- Distinctive: line 59
Test against a real data source (not mocked/fabricated output).
- Distinctive: line 71
Test against real data.
- Distinctive: line 78
Create documentation with real-world recipes.
- Distinctive: line 85
Add a "Chaining Recipes" section showing multi-skill workflows.
- Distinctive: line 86
Add a "What's Coming Next" section: blocked skills + their ticket links + owners.
- Distinctive: line 90
Data freshness: All examples MUST use real data from the current session.
- Distinctive: line 95
Tracking page: Add build phase rows. Add "What's Blocked" with ticket details.
- Distinctive: line 96
Repo: All skills committed and pushed.
- Very specific: line 103
250-line cap for skill.md: If a skill exceeds 250 lines, split into skill.md (core, <=250L) + examples.md + references/*.md.
- Distinctive: line 109
Do NOT report "Domain Build Complete" until ALL items are checked.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
Domain Build -- Build, Deploy, Document
Purpose: Execute the build/deploy/document cycle for a designed skill domain. Prerequisite: A completed
domain-designoutput -- prioritized skill roster, architecture decision, infrastructure map.
Announce at start: "I'm using the domain-build skill to implement the designed domain."
TODO Persistence
Use todo-management skill for all 4 phases. Tag tasks #plan-build-{domain}. Track each phase + post-build gates.
The 4-Phase Process
Input from domain-design: P0-P3 roster + architecture decision + walking skeleton design.
Phase 1: WALKING SKELETON
Build the simplest P0 skill to prove the full pattern end-to-end.
- Create
skills/{domain}/{skill-name}/skill.mdwith frontmatter:--- name: skill-name source: superpowers-plus triggers: [...] description: Use when... --- - Include the verified query/API call from the domain design.
- Define the output format the agent should produce.
- Install/deploy the skill per your overlay's install mechanism.
- Verify the skill is discoverable: run the discovery command for your skill system.
- Test against a real data source (not mocked/fabricated output).
- Commit and push.
Gate: Walking skeleton must be discoverable AND return real data before proceeding.
Phase 2: REMAINING P0s
For each remaining P0 skill in the roster:
- Create
skills/{domain}/{skill-name}/skill.md. - Deploy: run the install mechanism.
- Verify discoverable.
- Test against real data.
- Commit and push.
Batch commits are acceptable for remaining P0s.
Phase 3: COOKBOOK
Create documentation with real-world recipes. For each P0 skill, write one recipe with:
- When to use -- scenario description
- Ask the agent -- exact prompt to copy-paste
- What you get back -- real output (run the actual query, do not fabricate)
- Behind the scenes -- the query/API call
Add a "Chaining Recipes" section showing multi-skill workflows. Add a "What's Coming Next" section: blocked skills + their ticket links + owners.
Publish to your team's documentation platform nested under the design document.
Data freshness: All examples MUST use real data from the current session.
Phase 4: HANDOFF
- Design document: Mark P0 skills as Shipped with dates. Update file layout. Verify all ticket links.
- Tracking page: Add build phase rows. Add "What's Blocked" with ticket details. Update Key Findings.
- Repo: All skills committed and pushed. Skills discoverable via your install system.
- Report to user: Table of skills + status, blockers with tickets + owners, documentation links.
Hard Limits
250-line cap for skill.md: If a skill exceeds 250 lines, split into skill.md (core, <=250L) + examples.md + references/*.md. Verify: wc -l skills/{domain}/{skill-name}/skill.md.
Completion Checklist
Do NOT report "Domain Build Complete" until ALL items are checked.
Per-Skill Gates
skill.mdcreated with valid YAML frontmatterskill.mdis under 250 lines (wc -l)- Install/deploy exits 0
- Skill is discoverable via the skill system
- Tested against real data (no fabricated output)
- Committed and pushed
Documentation
- Design document -- P0 skills marked Shipped with dates, file layout updated
- Tracking page -- build phase rows added, blockers current
- Cookbook -- published with real data from the current session
- Skills index (if applicable) -- new skills listed with trigger conditions
- Skill count -- any file that states a total skill count is updated
Failure Modes
| Mode | Symptom | Recovery |
| Fabricated test output | Skill "works" in demo but fails in production | Always run against real data; never mock |
| Skipped validation | Build fails silently | Verify discovery + real-data test before claiming done |
| Cookbook uses stale data | Recipes mislead users | Pull fresh data during the current session |
| Incomplete checklist | "Done" declared prematurely | The checklist is a hard gate; partial completion = not done |
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.

