Skip to content
Back to skills

Ship Spec

ASecurity

Integrate and close an evidenced spec.

  • 14 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 20, 2026
ai-agentsgonodeapisecurity

Works with

  • api

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned October 5, 2026

npx -y skills add AIDDbot/AIDDbot --skill ship-spec --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Ship Spec?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Ship Spec
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/aiddbot-ship-spec/badge)](https://www.skillsdirectory.com/skills/aiddbot-ship-spec)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: ship-spec
description: Integrate and close an evidenced spec.
metadata:
  aiddbot-kind: primitive
user-invocable: true
---
# ship-spec

Your goal is to integrate one evidenced spec and close it with current schema and debt records.

Run `node .agents/aidd/aidd.mjs spec show` from the spec branch and stop on any listed blocker. Inspect each non-green report to confirm it contains only current findings.

Apply the spec's conceptual changes to `model.schema.md`, but update each affected `{project}.db.schema.md` and `{project}.api.schema.md` from the merged migrations, ORM schema, and routes, never from the spec. Touch a timestamp only when content changes.

Without running quality tools, reconcile the debt register through `node .agents/aidd/aidd.mjs debt` alone. Remove with `debt remove <id>` each D ID the spec's `Technical debt` section cites once verification is green, unless the spec's latest quality run still shows it; the next scan records again whatever remains. Then add every finding of a non-green qualification, `high` when it is classified `blocking`, and every failure or finding still present in a red report that reached revision 3, with `debt add "<title>" <high|medium|low> "<evidence>"`, unless an open item already describes it. Priority is `high` when it breaks behavior, security, or data; `medium` when it slows or complicates change; `low` otherwise. Never edit `control.json`, `debt.json`, or anything under `.aiddbot/` by hand: when a command refuses, fix what its error names and run it again.

Promote each durable, project-specific lesson that no automated check enforces into the project-rules table of the applicable `{source_root}/AGENTS.md`, as one short row with its scope and evidence-based reason, so the file stays short. Never record a one-off incident, a product requirement, or tool output. Add every helper the spec created in `shared` to the shared-primitives table of that `AGENTS.md`, with its contract and path.

Then run `node .agents/aidd/aidd.mjs run format` and commit whatever it rewrote with `node .agents/aidd/aidd.mjs commit "style: format"`; it is cosmetic, so a failing or unavailable format never stops shipping. Then run `node .agents/aidd/aidd.mjs release` from the spec branch, adding `--major` only when the delivery breaks compatibility for its users. It enforces the gate, derives the version from the spec type, writes it into the version files and `CHANGELOG.md`, marks the spec `shipped`, regenerates the spec index, commits, merges into the default branch, tags, and deletes the spec branch. Never edit a version or the changelog yourself. When it refuses, report its error and stop with the spec unshipped: never merge, tag, or integrate by other means, because only `release` writes the changelog and the spec index.

The result is one shipped spec, listed in the spec index, with current schema and debt records.

The core commits as `chore(release): {version}`.

Files in this skill

  • SKILL.md1.2 KB
  • assets/CHANGELOG.template.md143 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.