agentleFS
MageByte-ZeroMageByte-Zero/spec-superflow/skills/workflow-start/SKILL.md

workflow-start

Skill875 starsChanged 4 months agoMITView on GitHub

Primary entry point for the spec-superflow state-machine workflow. Invoke only when the user explicitly requests spec-superflow or an active change contains .spec-superflow.yaml and the user asks to continue that change. Generic proposal, specs, design, task, or contract files are not activation signals.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. ovargas/virtual-team★ 14same 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 repository11 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 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.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Before executing a CLI line below, replace its leading SSF with node "<plugin-root>/scripts/spec-superflow.mjs".

    line 10
  2. Very specific:

    Use only for an explicit spec-superflow request or an existing change.

    line 16
  3. Very specific:

    Update checks are optional (SSF runtime check-update), cached and non-blocking.

    line 28

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Workflow Start

Bundled runtime

Before executing a CLI line below, replace its leading SSF with node "<plugin-root>/scripts/spec-superflow.mjs"; <plugin-root> is the absolute directory two levels above this file. Never run SSF literally or call an ssf from PATH.

Workflow routing

A new request follows this file without loading legacy instructions. For an existing change, inspect the recorded state and receipt/plan (available through SSF workflow show <change-dir> --json and SSF execution show <dir> --json when needed). Current paths are workflow_variant: direct with a valid workflow-selection receipt schema_version: 3, or workflow_variant: planned with an execution plan schema_version: 2. Older receipts (schema 1/2), legacy plans and other recorded variants use the matching section of SSF runtime asset read docs/legacy-workflow.md; follow that section for this change. A direct variant alone is insufficient: legacy Quick/Hotfix/Lightweight can also record it. The internal workflow: full value alone is not a legacy signal. Missing or corrupt state, receipt or a required plan is a recovery problem; preserve evidence rather than reinitialize or downgrade the change.

Use only for an explicit spec-superflow request or an existing change. Ordinary coding does not require a spec workflow.

Resume

Run SSF resume <change-dir> --json once. It combines state, handoffs, checkpoints and current-plan inspection. Use the returned absolute change.path for all subsequent operations: recovery may redirect a stale source copy into its recorded worktree. Follow next_action; do not repeat all component commands unless diagnosing that result. A blocked continuation is not a retry instruction; resolve its named cause before running it again. Read only the artifacts relevant to the next action.

  • debugging: diagnose first, even if a plan is stale. Never dispatch an eligible wave during diagnosis.
  • closing: logical completion. If a recorded isolation still has physical finish pending, route to release-archivist for that remaining action, subject to existing merge authorization. Otherwise stop.
  • abandoned: stop.
  • A ready handoff needs review through SSF handoff resolve; an active handoff is non-blocking. Stale checkpoints are history, not current evidence.
  • A missing/stale planned execution plan blocks implementation, not investigation. Direct changes do not require a plan.

Update checks are optional (SSF runtime check-update), cached and non-blocking. Never make network access a recovery prerequisite.

New request

Infer scope and risks from the request and repository; do not turn CLI fields into a questionnaire. Validate the change name as a single safe relative path segment under changes/ and create only that directory. Reuse authorization already given. Never infer approval from artifact existence.

Offer only two paths, without a mode-selection interview:

  • Direct: clear requested change with a bounded proof. In the same turn run SSF workflow start <dir> --path direct --scope "<requested outcome and bounds>". No planning pack, recommendation receipt or execution plan is needed. Use affected tests or checks; --verification can record tdd/new-test/bounded when relevant. Task/file counts are advisory, not hard limits.
  • Planned: uncertain behavior, architecture or a user-requested plan. Draft proposal.md (scope, acceptance, risks) and tasks.md (ordered tasks and proof) together. Add specs/design only when their information is needed. Check shared interfaces against actual code once. Present one concrete plan for approval, reusing approval already covering it. Then run SSF workflow start <dir> --path planned --confirm --reason "<existing approval>". The CLI derives the execution plan and enters executing; no handwritten contract or intermediate approval sequence is required. Native + final is the default; add --mode sdd only for explicit delegation.

Continue to build-executor without another question. Ordinary debugging stays in executing. Scope changes update the affected planning sections and repeat planned start once with actual reapproval; do not traverse intermediate states. Nonsemantic corrections can use execution resync with a recorded reason, retaining review history.

Completion uses SSF workflow complete <dir> --verification-command "<required check command>". It runs the check once and preserves failures. Planned work also requires its current review and completed tasks; existing delta specs require synchronization. A user can explicitly accept known risks with workflow complete --accept-risk --confirm --reason "<decision and remaining issues>"; this records accepted-risk, never pass, and does not authorize integration.

Continue authorized internal work without phase-by-phase handoff questions. For all paths, read only the current task's source and evidence. Progress updates use one short paragraph; ask only for a new material decision, never “continue?” for authorized work.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.