gsd-orchestrator

Build software products autonomously via GSD headless mode. Handles the full lifecycle: write a spec, launch a build, poll for completion, handle blockers, track costs, and verify the result. Use when asked to "build something", "create a project", "run gsd", "check build status", or any task that requires autonomous software development via subprocess.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

80%

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

The body is highly actionable and token-efficient, with strong command/flag/exit-code reference material. Its main weakness is that the routing section references three workflow files that are absent from the bundle, breaking progressive disclosure and leaving some validation detail unreachable.

Suggestions

Add the missing workflow files (workflows/build-from-spec.md, monitor-and-poll.md, step-by-step.md) or remove those routing entries to avoid broken navigation.

Inline a brief validation/checkpoint step for risky operations (e.g. exit 1 -> inspect stderr and retry, exit 10 -> steer) so workflow_clarity does not depend solely on the absent workflow files.

Verify every path mentioned in the <routing> block resolves to a real bundle file before shipping.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence: dense flag/exit-code tables and copy-paste commands with no 'what is a subprocess' padding. Not below 3 because nearly every token earns its place; the narrative mental_model is minimal and load-bearing.	3 / 3
Actionability	Provides concrete, executable bash commands (e.g. 'gsd headless --output-format json new-milestone --auto 2>/dev/null') plus full flag and command tables. Not below 3 since examples are complete and copy-paste ready rather than pseudocode.	3 / 3
Workflow Clarity	A clear 5-step mental model and exit-code-to-action mapping exist, but the detailed sequenced workflows with validation checkpoints are routed to workflows/*.md files that do not exist on disk. Not 1 because the inline sequence and exit-code feedback are present; not 3 because the destructively-relevant validation/recovery loops are largely deferred to missing files, leaving checkpoint gaps.	2 / 3
Progressive Disclosure	The routing block cleanly signals one-level-deep references to references/*.md (which exist), but it also points to workflows/build-from-spec.md, monitor-and-poll.md, and step-by-step.md, none of which are present. Not 1 because organization and the existing references are solid; not 3 because broken workflow links undermine the disclosed navigation.	2 / 3
	Total	10 / 12 Passed

Description

100%

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 is concise, third-person, and explicitly covers both capabilities and natural trigger phrases. It avoids buzzwords and over-claims while clearly delimiting a distinct niche.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions ('write a spec, launch a build, poll for completion, handle blockers, track costs, and verify the result') rather than vague language. Not below 3 since the enumerated lifecycle steps are specific and concrete.	3 / 3
Completeness	Explicitly answers both what (full lifecycle actions) and when via a 'Use when asked to...' clause with concrete triggers. Not below 3 since both halves are clearly present and explicit.	3 / 3
Trigger Term Quality	Includes natural phrases users would say ('build something', 'create a project', 'run gsd', 'check build status'). Not below 3 because coverage spans both colloquial and product-name triggers.	3 / 3
Distinctiveness Conflict Risk	Scoped to a distinct niche (autonomous software dev via the GSD subprocess) with product-specific triggers unlikely to collide with other skills. Not below 3 because the GSD/subprocess framing is clearly distinct.	3 / 3
	Total	12 / 12 Passed

Validation

87%

Warnings & errors only

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

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: gsd-build/gsd-2
Path: gsd-orchestrator/SKILL.md
Commit: 33c00aa

Reviewed: about 14 hours ago

Table of Contents

Discovery Implementation Validation

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.