autonomous-investigation

The protocol behind every investigation skill. Use when AI research must proceed without you: search-plan gate, Fact/Inference/Assumption labels, confidence stacking, diffable outputs.

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/autonomous-investigation/SKILL.md

Quality

Content

71%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 well-crafted protocol skill with concrete, sequenced, actionable guidance and efficient prose that mostly assumes Claude's competence. Its chief weakness is progressive disclosure: the body signals several referenced files (template.md and two examples) and sibling skills that are not present in the bundle, leaving the navigation structure partly broken.

Suggestions

Ship the referenced bundle files (template.md, examples/protocol-in-action.md, examples/protocol-in-action-industrial.md) or remove the links so the one-level-deep navigation actually resolves.

Consider inlining a concrete numbered stable-output schema template in the Application section, since the protocol mandates a fixed schema but never shows one.

Trim the repeated "*Why it teaches:*" rationales and aphoristic one-liners to tighten token efficiency without losing the core contract.

Dimension	Reasoning	Score
Conciseness	The body is largely efficient and avoids explaining concepts Claude already knows (no "what a PDF is" filler); it assumes competence and earns most of its tokens, though the repeated "Why it teaches:" asides and aphoristic prose ("Research without a decision is a hobby", "an opinion wearing a badge") are minor trim opportunities that keep it just below the lean 5 anchor.	4 / 5
Actionability	Concrete, specific guidance is present throughout: an 8-step agent sequence, a 5-step implementer sequence, precisely defined Fact/Inference/Assumption labels, and a copy-paste worked example; as an instruction-only protocol it has no code, but the guidance is actionable, with only minor gaps such as the stable output schema being described rather than given inline.	4 / 5
Workflow Clarity	Multi-step processes are clearly sequenced with explicit checkpoints — the search-plan gate ("Continue unless the user revises it"), the question budget, and the gaps list all act as review gates; it stops short of the 5 anchor because there is no full validate→fix→retry feedback loop of the kind the top anchor calls for.	4 / 5
Progressive Disclosure	The SKILL.md itself is well-structured with clear sections and clearly signaled one-level-deep references (template.md, examples/protocol-in-action.md, examples/protocol-in-action-industrial.md, sibling skills), but none of the referenced bundle files actually exist in the bundle and no references/scripts/assets directories are present, so the progressive-disclosure promise of navigation to real detail is not fulfilled.	3 / 5
	Total	15 / 20 Passed

Description

75%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, concise description that answers both what and when with a concrete trigger phrase and named protocol mechanisms. Its main limitation is that the "what" describes a protocol's components rather than concrete user-facing actions, and a couple of terms lean on internal jargon.

Dimension	Reasoning	Score
Specificity	The description names several concrete protocol mechanisms — "search-plan gate, Fact/Inference/Assumption labels, confidence stacking, diffable outputs" — which are specific named components rather than vague abstractions, though they describe a protocol's parts rather than executable actions, keeping it just below the comprehensive-action anchor of 5.	4 / 5
Completeness	It states both what ("The protocol behind every investigation skill" with enumerated components) and when ("Use when AI research must proceed without you"), satisfying both halves; the "what" is somewhat abstract (a protocol rather than concrete actions), so it sits at 4 rather than the fully-concrete 5 anchor.	4 / 5
Trigger Term Quality	"Use when AI research must proceed without you" is a natural phrase a user would say, and "investigation skill" is a recognizable term; however terms like "confidence stacking" and "diffable outputs" are protocol jargon rather than common user vocabulary, so a few natural synonyms are missing.	4 / 5
Distinctiveness Conflict Risk	The niche is clear — autonomous, unattended research distinct from the sibling workshop-facilitation protocol — with minimal overlap risk; minor overlap remains with other research-oriented skills, placing it just below the minimal-conflict 5 anchor.	4 / 5
	Total	16 / 20 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
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 3 missing, 2 suspicious	Warning

	Total	14 / 16 Passed

Repository: deanpeters/Product-Manager-Skills
Path: skills/autonomous-investigation/SKILL.md
Commit: 1b5a524

Reviewed: 4 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.