migrate
Skill4 starsChanged 46 days agoMITView on GitHub
Migrates a project's documents (roadmap, stories, ADRs, specs, sprints) to the studio's current templates, keeping content, IDs and history; detects each format and shows the mapping and a rendered dry run before any write. Use after /adopt when documents predate the templates, when /update reports template drift, or for 'convert the roadmap', 'old story format', 'bring the ADRs up to date'.
63 other projects use a version of this.
Adopted by
Most recently taken up by
- tuist/agent-plugintheir own version, first seen Nov 8, 2026
- dsifry/metaswarmtheir own version, first seen Nov 8, 2026
- codewithmukesh/dotnet-claude-kittheir own version, first seen Nov 8, 2026
- jwm-axoni/claude-workbenchtheir own version, first seen Nov 8, 2026
- AsiaOstrich/EngramGraphtheir own version, first seen Nov 8, 2026
- DawnEver/claude-code-configtheir own version, first seen Nov 6, 2026
Used with
Projects with it also use sdd, scaffold and 6 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
liatrio-labsSkillsdd
Execute the Liatrio Spec-Driven Development (SDD) workflow when explicitly invoked by the user. NOTE: this skill is NOT intended to be dynamically loaded or automatically triggered; it should only ever be explicitly called by the user.
Also in 4 projects that use this
BerryKuipersSkillscaffold
Detect project stack, apply architecture patterns, wire quality gates, and scaffold features. Use when bootstrapping a project, adding a vertical slice, wiring CI/CD, adding Docker compose, or setting up quality gates.
Also in 4 projects that use this
agenticnotetakingSkillupgrade
Apply plugin knowledge base updates to an existing generated system. Consults the Ars Contexta research graph for methodology improvements, proposes skill upgrades with research justification. Never auto-implements. Triggers on "/upgrade", "upgrade skills", "check for improvements", "update methodology".
Also in 3 projects that use this
genomewalkerSkillcheckpoint
Save a checkpoint of task status, decisions, active files, and next steps
Also in 5 projects that use this
PydanticSkillcoverage
Fetch coverage diff from Codecov for the current branch or a specific PR. Shows uncovered lines, patch coverage, and overall coverage change.
Also in 3 projects that use this
JohnJBorenSkillci-cd
CI/CD pipeline design, GitHub Actions, testing automation, and deployment strategies
Also in 3 projects that use this
Other versions63 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 63
- garrytan/gbrain★ 32ksame name
- tuist/tuist★ 5.9ksame name
- parcadei/Continuous-Claude★ 3.9ksame name
- parcadei/Continuous-Claude-v2★ 3.9ksame name
- parcadei/Continuous-Claude-v3★ 3.9ksame name
- codewithmukesh/dotnet-claude-kit★ 808same name
- dsifry/metaswarm★ 426same name
- pipecat-ai/gradient-bang★ 420same name
- govctl-org/govctl★ 276same name
- AsiaOstrich/universal-dev-standards★ 84same name
53 more not listed
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 repository52 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- testing/CLAUDE.mdCLAUDE.md
- a11y-auditSkill
- adoptSkill
- api-contractSkill
- architecture-decisionSkill
- architecture-reviewSkill
- backlogSkill
- brainstormSkill
- changelogSkill
- code-reviewSkill
- collaboration-protocolSkill
- create-storiesSkill
- data-modelSkill
- dependency-auditSkill
- deploySkill
- design-systemSkill
- dev-storySkill
- docsSkill
- feature-specSkill
- game-conceptSkill
- hardenSkill
- helpSkill
- hotfixSkill
- impactSkill
- incidentSkill
- initSkill
- pentestSkill
- perf-auditSkill
- product-specSkill
- qa-planSkill
- refactorSkill
- release-checklistSkill
- retrospectiveSkill
- security-auditSkill
- setup-stackSkill
- skill-improveSkill
- skill-testSkill
- sprint-planSkill
- sprint-statusSkill
- stack-updateSkill
- startSkill
- story-doneSkill
- team-featureSkill
- team-gameSkill
- team-releaseSkill
- team-securitySkill
- tech-debtSkill
- test-setupSkill
- threat-modelSkill
- updateSkill
- ux-specSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 50 edited versions we read keep these.
2
Distinctive
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
- Distinctive: line 31
Show the exact rules per document type: before anything is rendered:
- Very specific: line 33
every task line becomes - [ ] ID · Title + markers in the legend's order.
- Very specific: line 59
After the "write" answer: touch .claude/.write-consent (rule 7 — the consent-guard hook checks the marker).
- Distinctive: line 62
When HEAD is a story branch, name it and ask: switch to the default branch and commit there (Recommended — migrated documents are …
- Very specific: line 63
Before asking, record the gate — <hooks>session-state.sh set Gate "/migrate Phase 4: commit <type>?" (<hooks> is .claude/hooks/ in copy …
“Common” is measured: kept by at least 6 in 10 of the 50 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
Migrate — documents to the current templates
Reply in the project conversation language (CLAUDE.md → Language); code, identifiers, paths and commit messages stay in English.
Templates .claude/docs/templates/{roadmap,story,adr,product-spec,feature-spec,sprint-plan,backlog,decisions}.md;
rules .claude/rules/docs-format.md (required sections per document type). Writes only after "May I write?", one
document type per gate. Not a rewrite: the content, the decisions and every ID survive; only structure,
markers and links change. /adopt classifies the gap, this skill closes it.
Phase 1: Detect
- Initialised? No
.claude/docs/workflow-catalog.yaml→BLOCKED (not initialised — run /init), and stop. - Files in scope. For each type in the argument (
all= every type), list the files that exist. No files of the requested type at all →BLOCKED (no documents of this type). - Fingerprint each file:
- roadmap: the
roadmap-format:header comment (v3.1), a reference-style## Linksblock,📅 sprint-NNmarkers, numbered "Owner decisions", or a foreign format (a companion tool's table, a plain list, a spreadsheet export); - stories: the template sections present (Goal, Acceptance criteria as a table, Definition of Done);
- ADRs: Status / Context / Options / Decision / Consequences / Verification;
- specs: the numbered template sections;
- sprints: Goal / Stories / Dependency updates / Risks / Retrospective.
- roadmap: the
- Keep the documents' prose language. Compare headings by structure and count; never translate them back to English.
- Show the table "document → format found → drift (sections missing, markers to convert) → action". Documents already in the current format are listed as
up to dateand never touched. Every document up to date →COMPLETE (0 documents migrated, N up to date), with no write gate.
Phase 2: Mapping
- Show the exact rules per document type before anything is rendered:
- Roadmap → v3.1:
- every task line becomes
- [ ] [ID](path) · Title+ markers in the legend's order; - IDs are kept verbatim (a foreign scheme such as
T-12orH-37stays — the format allows any stable prefix); - paths become file-relative inline links (a card that does not exist yet links to
backlog.md#id, or stays a bare ID with a note); 📅 sprint-NN→ the sprint heading with ISO dates;- "Owner decisions" →
production/decisions.md(D-NN), referenced from the lines that wait on them with⛔ [D-NN](decisions.md#d-nn); - ideas without a card →
production/backlog.md(I-NNN); - closed sprints, the Backlog and the Legend fold into
<details>; - the three-line header names the manager, the source of truth and today's date;
- the
## Docsboard is generated from the actual files.
- every task line becomes
- Stories: missing sections are added with
n/a — reasonor filled from the card's prose (criteria in prose → the Given/When/Then table, one row per criterion, test column[to map]); status and links kept. - ADRs:
/architecture-decision retrofitsemantics applied in bulk — an option list of at least two (the rejected alternative reconstructed from the text when it is there, otherwise[not recorded]), Verification added as[to define]; Status kept. - Specs: sections renumbered to the template, missing ones
n/a — reason; acceptance criteria never dropped. - Sprints: the
## Retrospectiveand## Dependency updatessections added empty where missing; the header getsStatus:—closed <date>(the date of the retrospective commit, else of the last⏱/🔗story) when the sprint's roadmap block is folded or every one of its lines is[x],activeotherwise;## Storiesrows take their status from the roadmap lines (the file's own column drifts as soon as a story closes without/story-donewriting it).
- Roadmap → v3.1:
- Questions before the render. Anything that cannot be mapped mechanically is listed and asked as one
AskUserQuestionwith the real alternatives. Never guess it.
Phase 3: Dry run
- Render each converted document in the chat message (rule 7: a roadmap as its list, a criteria table as a table).
- Summary line per file:
N lines → M lines · K IDs kept · sections added: … · questions: 0. - Compare IDs and links before and after. Every ID of the original must still be there:
grep -oE '\b[A-Z]{1,4}-[0-9]+\b' <file> | sort -uon the original and on the converted files; name any missing ID and fix the conversion before the write gate. IDs the migration introduces (D-NN,I-NNN) are listed separately. --dry-runstops here with the verdictDRY RUN (N documents ready, M questions).
Phase 4: Write (one type per gate)
- Write gate, one
AskUserQuestionper type: "May I write<the files of this type>(N files; git history is the backup)?" — write (Recommended) · show the full diff first · skip this type. One answer never covers a second type. - After the "write" answer:
touch .claude/.write-consent(rule 7 — the consent-guard hook checks the marker). Then write the files. - Commit gate for the batch (rule 7,
.claude/docs/git-workflow.md§ Documents), right after the write, oneAskUserQuestion:docs: migrate <type> to template <version>, staging exactly the written files of this type.- On the default branch when no story work is in progress.
- When HEAD is a story branch, name it and ask: switch to the default branch and commit there (Recommended — migrated documents are pipeline-wide) · commit here (the documents belong to this story) · leave uncommitted.
- Before asking, record the gate —
<hooks>session-state.sh set Gate "/migrate Phase 4: commit <type>?"(<hooks>is.claude/hooks/in copy mode and${CLAUDE_PLUGIN_ROOT}/hooks/in plugin mode) — and clear it after the answer (<hooks>session-state.sh set Gate "—"). - Nothing is committed without the answer; one commit per type, never one answer covering two types.
- After the roadmap:
/helpmust be able to read it. Run the read/helpPhase 1 step 3 performs on the roadmap (the first open- [ ]line, plus the count of open lines) in this skill and show the result —/helpitself is the hand-off below, not a call from here (this skill does not listSkill).
Verdict: COMPLETE (N documents migrated, M up to date) | DRY RUN (…) | BLOCKED (not initialised — run /init |
no documents of this type). Next step — one AskUserQuestion: /help (Recommended) · migrate the next type ·
stop here.
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.

