agentleFS
nishant-tamilselvannishant-tamilselvan/AEGIS/.github/skills/artifact-management/SKILL.md

artifact-management

Skill1 starsChanged 48 days agoMITView on GitHub

Create, update, and validate the living artifact documents — the six business artifacts (product requirements, functional requirements, NFRs, user journey map, system blueprint, executive briefing) and the five technical artifacts (interface specifications, data/security/deployment/observability architecture). Use when scaffolding artifacts, adding or changing requirements or design, running consistency validation, or bumping document versions.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 5 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After any artifact edit, to validate consistency before ending a phase.

    line 26
  2. Very specific:

    Never hand-write a document shape.

    line 30
  3. Very specific:

    Every application lives in its own folder: docs/artifacts/<app-name>/.

    line 32
  4. Distinctive:

    After every change, run validation and resolve all errors before ending the phase.

    line 39
  5. Very specific:

    If the application is brand new, also add a row for it in docs/artifacts/README.md.

    line 46
  6. Distinctive:

    Add or modify rows in the relevant table(s), assigning the next sequential id.

    line 81
  7. Distinctive:

    Keep each requirement atomic and testable.

    line 83
  8. Very specific:

    Add --strict to also fail on advisory warnings (e.g.

    line 101
  9. Distinctive:

    Every cross-referenced id resolves to an id defined in some artifact.

    line 110
  10. Very specific:

    Every functional requirement traces to a PR-* or UJ-* (warning, or error in --strict).

    line 111
  11. Very specific:

    Pick the app first: Every command targets docs/artifacts/<app-name>.

    line 118

≈ “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

Artifact Management

Authoritative procedure for producing and maintaining the workflow's living artifacts: six business artifacts (ideation phase) and five technical artifacts (architecture phase). Only the artifact-manager agent should apply these edits; other agents propose content. Architecture Decision Records are handled by the sibling adr-management skill.

Artifact Manager is also the sole writer of phase-3 state under docs/artifacts/<app>/implementation/, but those documents use the separate implementation-management procedure and CLI; they are not additional business or technical artifact types.

When to use

  • Starting a new project (scaffold the full set from templates).
  • A new requirement, persona, journey, or component was agreed during ideation.
  • An existing requirement changed, was refined, or removed.
  • After any artifact edit, to validate consistency before ending a phase.

Ground rules

  1. Never hand-write a document shape. Always start from the templates in ./assets/templates/.
  2. Every application lives in its own folder: docs/artifacts/<app-name>/. All commands below target that per-application directory — never bare docs/artifacts. The active <app-name> is provided by the orchestrator at session start.
  3. Requirement IDs are stable and never reused within an application. The same id may exist independently in another app. Follow ./references/id-conventions.md.
  4. Frontmatter must match ./references/artifact-schema.md.
  5. After every change, run validation and resolve all errors before ending the phase.

Procedure

1. Scaffold (first time only)

Replace <app> with the active application's folder (e.g. docs/artifacts/customer-portal). If the application is brand new, also add a row for it in docs/artifacts/README.md.

python -m artifact_tools scaffold product-requirements docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold functional-requirements docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold non-functional-requirements docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold user-journey-map docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold system-blueprint docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold executive-briefing docs/artifacts/<app> --project "<Name>"

Type aliases also work: prd, fr, nfr, journey, blueprint, briefing.

In the architecture phase, scaffold the technical artifacts the same way:

python -m artifact_tools scaffold interface-specifications docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold data-architecture docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold security-architecture docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold deployment-topology docs/artifacts/<app> --project "<Name>"
python -m artifact_tools scaffold observability-strategy docs/artifacts/<app> --project "<Name>"

Technical aliases: interfaces, data, security, deployment, observability.

Scaffolding interface-specifications also creates the <app>/interfaces/ contract store (synchronous/, asynchronous/, graphql/). That single-file document is a routing index — record real API/event/schema contracts as native .yaml / .json / .graphql / .proto files there, one contract per file, and link each from the index. Re-create the store at any time with python -m artifact_tools interfaces init docs/artifacts/<app>.

2. Edit content

  • Add or modify rows in the relevant table(s), assigning the next sequential id.
  • Fill cross-reference columns (traces_to, satisfies, mitigates) with existing ids.
  • Keep each requirement atomic and testable.
  • To retire an item, set its row Status to removed; do not delete the row.

3. Record the change and bump the version

python -m artifact_tools changelog docs/artifacts/<app>/functional-requirements.md "Added FR-014 for offline sync" --phase 1

This appends a changelog line and bumps version + last_updated in frontmatter.

4. Validate (always)

python -m artifact_tools validate docs/artifacts/<app>
  • Exit code 0 = consistent. Non-zero = errors that must be fixed.
  • Add --strict to also fail on advisory warnings (e.g. missing traceability).
  • Validate every application at once with python -m artifact_tools validate docs/artifacts; each app is checked in isolation and findings are prefixed with the app folder name.

What validation checks

  • Frontmatter completeness and valid status / version / last_updated / phase.
  • artifact field matches the file type.
  • Ids are unique within a table and use the correct prefix for the document.
  • Every cross-referenced id resolves to an id defined in some artifact.
  • Every functional requirement traces to a PR-* or UJ-* (warning, or error in --strict).

Multiple applications

A workspace can hold many applications, each in its own docs/artifacts/<app-name>/ folder with its own full artifact set, architecture-decisions/, and interfaces/.

  • Pick the app first. Every command targets docs/artifacts/<app-name>. Never write artifacts to the bare docs/artifacts root — that folder only holds the applications index (README.md).
  • New application? Choose a short kebab-case slug for the folder, scaffold into it, and add a row to docs/artifacts/README.md.
  • IDs are per application. FR-001 in one app is unrelated to FR-001 in another. Cross-references only resolve within the same application's folder.

Resources

  • Templates: ./assets/templates/
  • Schema: ./references/artifact-schema.md
  • ID conventions: ./references/id-conventions.md

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.