skillforge
Skill3 starsChanged 8 days agoMITView on GitHub
SUPERSEDED — skill authoring is now split three ways, and this skill duplicates all three without matching any of them: `skill-creator` (Anthropic plugin) owns measurement, since it carries the only executable eval harness in the set — trigger-rate testing on a held-out split, with-skill vs baseline runs, benchmark aggregation; `superpowers:writing-skills` owns wording form, including the match-the-form-to-the-failure model; and the studio's `skill-smith` agent owns house rules — catalog conformity, the frontmatter contract, upstream-PR discipline. Kept unchanged during the migration window; new work should load the successors. Its own conventions do not fit this catalog: the `.skills/<name>/` layout matches neither `.claude/skills/` nor `skills/`, its `.skills-registry.json` and `skdd forge` tooling are not wired here, and its "under 200 lines" is a third size rule alongside Anthropic's 500 lines and this catalog's CI-enforced byte budget. Use when you need the historical text; otherwise load a successor.
9 other projects use a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- hamzzaqureshi/gravidextheir own version, first seen Nov 8, 2026
- ngmeyer/skillstheir own version, first seen Nov 1, 2026
- andrewdorsey-glitch/Claude-Skillstheir own version, first seen Oct 8, 2026
- bmaltais/skillstheir own version, first seen Sep 6, 2026
- boshen99/skillforgetheir own version, first seen Aug 26, 2026
- tripleyak/SkillForgetheir own version, first seen Jul 29, 2026
Other versions9 projects use a version of this.
Word for word · 0
None found.
Edited · 1
- zakelfassi/skills-driven-development★ 1891% same
Wrote their own version · 8
- tripleyak/SkillForge★ 884same name
- mmlong818/skillforge★ 86same name
- boshen99/skillforge★ 4same name
- ngmeyer/skills★ 4same name
- bmaltais/skills★ 2same name
- yvzhou1111/skillforge★ 2same name
- hamzzaqureshi/gravidex★ 1same name
- andrewdorsey-glitch/Claude-Skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository66 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- clud-bug-collaborationSkill
- critical-issues-onlySkill
- evidence-based-reviewSkill
- respect-existing-conventionsSkill
- skill-frontmatter-qualitySkill
- apca-contrastSkill
- api-contract-enforcementSkill
- brand-voice-reviewSkill
- chroma-harmonizationSkill
- clud-bug-collaborationSkill
- component-sizing-principlesSkill
- component-sizingSkill
- composing-a-screenSkill
- consuming-a-design-systemSkill
- conventions-templateSkill
- critical-issues-onlySkill
- curating-a-skill-catalogSkill
- designing-a-design-systemSkill
- designing-elite-uiSkill
- design-system-consistencySkill
- design-token-namingSkill
- dtcg-formatSkill
- empirical-design-principlesSkill
- evidence-based-reviewSkill
- finding-a-catalog-skillSkill
- frontend-a11ySkill
- guarding-a-regressionSkill
- ios-safari-surfacesSkill
- line-height-gridSkill
- logmindSkill
- motion-continuitySkill
- oklch-color-spaceSkill
- orchestrating-agent-delegationSkill
- orchestrating-a-multi-agent-runSkill
- orchestrating-elite-agent-qaSkill
- palette-relationshipsSkill
- pii-and-complianceSkill
- proving-an-absenceSkill
- respect-existing-conventionsSkill
- retiring-a-superseded-decisionSkill
- reviewing-design-workSkill
- semver-design-tokensSkill
- session-heartbeatSkill
- skill-frontmatter-qualitySkill
- spacing-systemSkill
- technical-writingSkill
- test-disciplineSkill
- thrill-designSkill
- token-frugal-toolingSkill
- token-naming-conventionsSkill
- type-scaleSkill
- udts-component-sizing-laddersSkill
- udts-dtcg-extensionsSkill
- udts-linter-rulesSkill
- udts-naming-conventionSkill
- udts-reviewSkill
- udts-semver-defaultsSkill
- udts-spacing-defaultsSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
9 rules. 0 are common, 5 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 8 edited versions we read keep these.
5
Distinctive
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
- Distinctive: line 17
Create well-formed, spec-compliant skills from observed patterns.
- Distinctive: line 81
If the skill involves file generation or automation:
- Very specific: line 93
Update .skills-registry.md at the project root (same level as .skills/).
- Very specific: line 99
If the project uses the machine-readable registry (.skills-registry.json), update it too — skdd forge handles both formats automatically.
- Distinctive: line 103
When you use a skill and encounter something it doesn't cover:
- Very specific: line 105
Add the new edge case or step to the existing SKILL.md
- Distinctive: line 106
If the skill is getting too long (>200 lines), split it
- Very specific: line 107
Update last-used and increment usage-count in the registry
- Distinctive: line 111
Before committing a new skill:
“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
SkillForge
Create well-formed, spec-compliant skills from observed patterns.
When to Forge
✅ Forge when:
- You've done the same sequence 2-3 times in a session
- A project convention isn't documented anywhere
- You solved a hard problem with a reusable solution
- Someone asks you to create a skill
❌ Don't forge when:
- It's a one-time task
- An existing skill already covers it (update instead)
- The "skill" is just a single command (use a script alias)
Steps
1. Name the pattern
- What problem does this skill solve?
- What triggers it? (be specific — the
descriptionfield is the discovery surface) - What are the inputs and outputs?
2. Choose a name
kebab-case, 1-64 characters- Verb-led when possible:
deploy-preview,scaffold-component,triage-bug - One responsibility per skill
3. Create the skill directory
mkdir -p .skills/<skill-name>4. Write SKILL.md
Use this skeleton:
---
name: <skill-name>
description: <what it does>. Use when <triggers>.
metadata:
forged-by: <agent-id>
forged-from: <session-or-context>
forged-reason: "<why this was created>"
---
# <Skill Name>
## Inputs
- ...
## Steps
1. ...
2. ...
## Conventions
- Project-specific patterns that apply
## Edge Cases
- Known gotchas or special handling5. Add scripts (optional)
If the skill involves file generation or automation:
.skills/<skill-name>/
├── SKILL.md
├── scripts/
│ └── run.sh # Executable automation
└── references/
└── conventions.md # Detailed reference (keeps SKILL.md lean)6. Register the skill
Update .skills-registry.md at the project root (same level as .skills/). Create it if it doesn't exist:
| <skill-name> | local | <today> | 1 | <description> |If the project uses the machine-readable registry (.skills-registry.json), update it too — skdd forge handles both formats automatically.
Updating an Existing Skill
When you use a skill and encounter something it doesn't cover:
- Add the new edge case or step to the existing SKILL.md
- If the skill is getting too long (>200 lines), split it
- Update
last-usedand incrementusage-countin the registry
Quality Checklist
Before committing a new skill:
nameis kebab-case, ≤64 charsdescriptionincludes what it does AND when to use it- Steps are numbered and actionable
- No hardcoded paths, secrets, or environment-specific values
- SKILL.md is under 200 lines (move details to
references/) - Registered in
.skills-registry.md
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.

