cavecrew

Decision guide for delegating to caveman-style subagents. Tells the main thread WHEN to spawn `cavecrew-investigator` (locate code), `cavecrew-builder` (1-2 file edit), or `cavecrew-reviewer` (diff review) instead of doing the work inline or using vanilla `Explore`. Subagent output is caveman-compressed so the tool-result injected back into main context is ~60% smaller — main context lasts longer across long sessions. Trigger: "delegate to subagent", "use cavecrew", "spawn investigator/builder/reviewer", "save context", "compressed agent output".

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

92%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 tightly written, highly actionable delegation guide with concrete output contracts, clear chained workflows, and built-in validation feedback loops. Its only weakness is mild redundancy where the body re-explains the compression rationale already present in the description.

Suggestions

Trim or fold the 'Why this exists (the real win)' section: the ~60% compression claim and token example repeat the description's value prop — keep one concise sentence or move the rationale into the decision-table rule of thumb.

Consider moving the three detailed output-contract templates into a references/ file (e.g. OUTPUT_FORMATS.md) so the main body stays a decision overview, if the contracts grow beyond their current terse form.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean — terse tables, output-contract templates, and a 'What NOT to do' list that all earn their place — but the 'Why this exists (the real win)' section re-explains the ~60% compression benefit already stated in the description via a worked 2k-vs-700 token example, a minor instance of over-explanation that could be trimmed.	4 / 5
Actionability	Provides fully actionable guidance: exact output-contract formats with literal terminal tokens ('too-big.', 'needs-confirm.', 'regressed.'), a decision table mapping task types to agents, and concrete chaining patterns — copy-paste-ready templates covering the common cases.	5 / 5
Workflow Clarity	The 'Locate → fix → verify' chain is explicitly numbered and embeds validation/feedback loops (cavecrew-reviewer audits the diff; builder emits 'verified: re-read OK \| mismatch @ path:line' and 'regressed.' for error recovery), matching the anchor for clear sequence with explicit validation steps.	5 / 5
Progressive Disclosure	No bundle files exist and none are needed; the self-contained ~70-line body is organized into well-signaled sections (decision table, output contracts, chaining patterns, anti-patterns) with no nested references, so it reads as a clear overview appropriate for a single-purpose skill.	5 / 5
	Total	19 / 20 Passed

Description

88%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.

A strong, third-person description that names concrete capabilities and supplies explicit trigger guidance. It is highly specific and complete, with only minor overlap risk from a couple of broad trigger terms.

Dimension	Reasoning	Score
Specificity	Names the domain and three concrete agent actions — 'cavecrew-investigator (locate code)', 'cavecrew-builder (1-2 file edit)', 'cavecrew-reviewer (diff review)' — giving comprehensive coverage of capabilities rather than generic language.	5 / 5
Completeness	Clearly answers both 'what' (decision guide naming three presets and their jobs) and 'when' (an explicit 'Trigger:' clause with concrete trigger phrases), matching the top anchor.	5 / 5
Trigger Term Quality	Explicit Trigger clause lists natural phrases like 'delegate to subagent', 'use cavecrew', 'save context', but a couple of entries ('compressed agent output', 'spawn investigator/builder/reviewer') lean technical and omit common synonyms; good but not fully comprehensive coverage.	4 / 5
Distinctiveness Conflict Risk	The distinctive 'cavecrew' naming and three named presets carve a clear niche, but broad triggers like 'delegate to subagent' and 'save context' create minor overlap risk with general subagent/delegation skills, so it sits just below the top anchor.	4 / 5
	Total	18 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: JuliusBrussee/caveman
Path: skills/cavecrew/SKILL.md
Commit: 14d4f2e

Reviewed: 5 days 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.