agentleFS
tuana-vntuana-vn/system-reverse-engineer/.github/skills/artifact-detailed-design/SKILL.md

artifact-detailed-design

Skill0 starsChanged 2 months agoMITView on GitHub

Produce a source-backed detailed design for exactly one registered design artifact at a time, preserving requirement/TDD traceability, current-source anchors, runtime/data provenance, compatibility, and explicit proposed-vs-existing boundaries.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 7 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use test design when it clarifies verification obligations.

    line 34
  2. Distinctive:

    verify the Artifact ID exists in registry

    line 40
  3. Very specific:

    verify its requirement/gap/TDD mappings

    line 41
  4. Distinctive:

    check whether dependency designs are required first

    line 44
  5. Distinctive:

    If a prerequisite artifact is unresolved and blocks design:

    line 46
  6. Very specific:

    Never describe a proposed class/module/method as existing.

    line 60
  7. Distinctive:

    all exposure paths materially affected

    line 129
  8. Distinctive:

    Do not omit these because the artifact appears "internal".

    line 172
  9. Distinctive:

    Use the active workflow-declared DD output path under the current run root.

    line 205

≈ “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 Detailed Design

Purpose

Detailed design is generated per registered artifact, not as an unconstrained rewrite of the TDD.

One run handles exactly one Artifact ID.

Requirement
→ Gap
→ TDD Decision
→ Artifact Registry Entry
→ Detailed Design

Inputs

Required:

  • Artifact ID
  • design artifact registry
  • verified TDD
  • TDD verification report
  • requirement/gap artifacts
  • current repository

Use test design when it clarifies verification obligations.

1. Reverify Artifact Context

Before designing:

  • verify the Artifact ID exists in registry
  • verify its requirement/gap/TDD mappings
  • reverify existing source anchors
  • load direct dependencies
  • check whether dependency designs are required first

If a prerequisite artifact is unresolved and blocks design:

DD_BLOCKED_DEPENDENCY

2. Separate Existing / Proposed / Unknown

Use explicit sections:

CURRENT SOURCE-BACKED
PROPOSED
UNKNOWN / SOURCE GAP

Never describe a proposed class/module/method as existing.

3. Generic Core Design

For every artifact define:

  • responsibility
  • scope / non-goals
  • source anchors or proposed location
  • dependencies
  • public/internal contract
  • inputs/outputs/state
  • runtime behavior
  • error/failure behavior
  • configuration/lifecycle where applicable
  • compatibility constraints
  • concurrency/threading only when source/design makes it relevant
  • security/sensitive-data considerations only when relevant
  • verification obligations

4. Type-Specific Design

CLASS / INTERFACE / MODULE / COMPONENT

Include:

  • existing/proposed signatures
  • callers/consumers
  • dependencies
  • lifecycle/ownership
  • exceptions/results
  • before/after relationships

API CONTRACT

Include:

  • method/operation
  • request inputs
  • validation/combination rules
  • response/output
  • error contract
  • compatibility

DATA MODEL

Include:

  • fields/types
  • source/provenance
  • nullability/default
  • serialization/persistence
  • sensitive-data classification
  • compatibility

CONFIGURATION

Include:

  • source/default/override precedence
  • consumer
  • reload/lifecycle
  • invalid/missing behavior
  • backward compatibility

INTEGRATION

Include:

  • operation/request construction
  • routing/selection
  • result/error semantics
  • timeout/retry/fallback
  • data provenance
  • all exposure paths materially affected

RUNTIME FLOW / STATE MACHINE

Include:

  • triggers
  • states/branches
  • transitions
  • data/context propagation
  • side effects
  • success/failure paths
  • before/after Mermaid

PERSISTENCE

Include:

  • schema/object/key
  • read/write paths
  • transaction/consistency semantics
  • migration/backward compatibility

ERROR HANDLING

Include:

  • source error
  • translation
  • observable error
  • preservation/masking rules

Other types should use the same evidence-first principle and explain type-specific details.

5. Provenance and Exposure

When applicable include:

field/data provenance:
trigger → source → transform → mapping/state → output

boundary/exposure:
boundary ← all material invocation sites ← external/internal roots

Do not omit these because the artifact appears "internal".

6. Proposed Change Table

DD ItemExisting Source AnchorCurrent BehaviorProposed BehaviorRequirement/Gap/TDDRiskVerification

7. Detailed Design Gate

Before DD_READY_FOR_VERIFICATION:

[ ] artifact registry mappings preserved
[ ] existing source anchors reverified
[ ] proposed artifacts/methods clearly marked
[ ] runtime behavior is concrete enough to implement
[ ] field/data provenance covered where applicable
[ ] boundary/exposure covered where applicable
[ ] error/failure behavior defined
[ ] compatibility/non-goals preserved
[ ] dependencies identified
[ ] verification obligations identified
[ ] no HIGH/CRITICAL source-resolvable gap hidden as a design assumption

Statuses:

  • DD_READY_FOR_VERIFICATION
  • DD_BLOCKED_SOURCE_GAP
  • DD_BLOCKED_DEPENDENCY

Output

Use the active workflow-declared DD output path under the current run root.

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.