CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

gsd-ai-integration-phase

Generate an AI-SPEC.md design contract for phases that involve building AI systems.

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./pi/package-skills/gsd-ai-integration-phase/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A lean, well-structured orchestrator wrapper: it gives genuinely concrete dispatch mechanics with fail-closed error handling and a clear role sequence, while delegating all substantive workflow content to external engine files. Its main weaknesses are the unresolved `__GSD_PI_ENGINE_ROOT__` placeholder paths and the absence of the referenced files in the bundle, which make the skill non-self-contained and navigation-dependent on external state.

Suggestions

Explain or resolve the `__GSD_PI_ENGINE_ROOT__` placeholder (how to substitute it at invocation time) so the execution_context paths are directly readable.

Inline a brief summary of the workflow gates that must be preserved (or their locations in the referenced files) so the skill remains actionable if a reference fails to load.

Replace "Execute end-to-end. Preserve all workflow gates." with the concrete gate list or checkpoint commands, mirroring the specificity already present in the dispatch section.

DimensionReasoningScore

Conciseness

The body is lean and assumes competence: terse directives like "Execute end-to-end. Preserve all workflow gates." and a one-line flow summary, with no explanation of concepts Claude already knows. The dense 'Pi dispatch' paragraph has minor trimmable over-explanation (e.g., restating how an Agent(...) call maps to role/task), keeping it at anchor 4 rather than 5.

4 / 5

Actionability

Concrete, executable guidance is present: call gsd_dispatch with named `role` and `task` parameters, explicit fail-closed handling (STOP on unresolved role or failed child, optional vs. required step behavior), and a fully expanded placeholder path convention for reading execution-context files. Not 5 because the actual workflow content and commands live entirely in external files referenced via the unresolved `__GSD_PI_ENGINE_ROOT__` placeholder, and "Preserve all workflow gates" delegates rather than instructs.

4 / 5

Workflow Clarity

The sequence is explicit ("Orchestrates gsd-framework-selector → gsd-ai-researcher → gsd-domain-researcher → gsd-eval-planner" plus a labeled flow), and the dispatch section includes explicit error checkpoints (stop on unresolved role/failed child, offer skip for optional steps). Not 5 because validation gates are only referenced ("Preserve all workflow gates") without being enumerated or defined, and the step details are delegated to external files. No destructive/batch operation cap applies.

4 / 5

Progressive Disclosure

The body is a clean overview: clearly labeled sections (objective, execution_context, context, process) and three one-level-deep references grouped under a dedicated, well-signaled block. Not 5 because the referenced paths use the unexpanded `__GSD_PI_ENGINE_ROOT__` placeholder and none of the referenced files exist in the skill's bundle (no references/, scripts/, or assets/ directories), so navigation depends entirely on external resolution the skill doesn't explain.

4 / 5

Total

16

/

20

Passed

Description

50%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description states a clear, single-action 'what' with a concrete artifact name (AI-SPEC.md), but it lacks an explicit 'Use when...' trigger clause, offers thin natural-language keyword coverage, and doesn't convey the orchestration steps the skill performs. It sits squarely at the midpoint of the rubric.

Suggestions

Add an explicit trigger clause, e.g. 'Use when a GSD phase involves building AI features, LLM integration, or model-powered functionality.'

List the concrete actions performed (select an AI framework, run research, design an eval strategy) so the 'what' is comprehensive rather than a single verb.

Include natural user phrasings/synonyms (AI feature, model integration, eval planning) to improve trigger-term coverage and reduce overlap with other AI skills.

DimensionReasoningScore

Specificity

"Generate an AI-SPEC.md design contract" names one concrete action and a concrete artifact, and "for phases that involve building AI systems" names the domain, but coverage is not comprehensive — no mention of the underlying actions (framework selection, research, eval planning) that the skill actually performs. It matches the 1-2-concrete-actions anchor and falls short of anchor 4's 'several specific actions'.

3 / 5

Completeness

The 'what' is clear (generate an AI-SPEC.md design contract), but there is no explicit 'Use when...' trigger clause — "for phases that involve building AI systems" only weakly implies the 'when', so completeness is capped at 3 per the judging guideline. Not 4 because the 'when' is not explicitly stated as trigger guidance; not 2 because the 'what' is clear rather than vague.

3 / 5

Trigger Term Quality

Terms like "AI systems", "phases", and "design contract" are somewhat relevant, but the phrasing is workflow-jargon-heavy ("AI-SPEC.md", "design contract") and misses natural variations users would say (AI feature, LLM integration, model integration, AI eval). Anchor 3 ('some relevant keywords but missing common variations or synonyms') fits; not 2 because it is more than one or two generic keywords, not 4 because coverage of natural terms is thin.

3 / 5

Distinctiveness Conflict Risk

"Generate an AI-SPEC.md design contract" carves out a fairly distinct artifact niche, but "for phases that involve building AI systems" is broad and would overlap with other AI-related skills (research, eval planning, AI implementation). Anchor 3 ('somewhat specific but could still overlap with similar skills') fits best; not 4 because the AI-systems scope is broad rather than a clear niche with distinct triggers.

3 / 5

Total

12

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

open-gsd/gsd-core

Path

pi/package-skills/gsd-ai-integration-phase/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.