investigate

Analyzes observability data — logs, traces, errors, sessions, and metrics — to find root cause and actionable evidence. Use when the user reports a bug, an unexpected behavior, or asks about patterns across application data.

Quality

90%

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

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.

An exceptionally tight, actionable skill body with a well-sequenced workflow and concrete tool-level guidance. Its one real weakness is that the progressive-disclosure references point to companion files that are not present in the bundle.

Suggestions

Create the referenced companion files (logs.md, traces.md, errors.md, sessions.md, metrics.md) under references/ so the promised one-level-deep navigation actually resolves.

Alternatively, if a companion file is not yet written, mark it as such inline (e.g., "(coming soon)") or remove the reference rather than shipping a broken link.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence: no explanations of what logs, traces, or sessions are, and every bullet delivers concrete guidance; the brief framing paragraph ("Real investigations almost always touch more than one product") earns its place by establishing the cross-product workflow.	5 / 5
Actionability	Names exact MCP tools, a specific 50-entry limit, concrete field-name variations ("spanName vs span_name"), and an exact remediation recipe (config change, flag flip, code fix with file/line, specific query to run); per the scoring note, absence of code is not penalized for an instruction-only skill when guidance is this actionable.	5 / 5
Workflow Clarity	A clear 7-step numbered sequence with a feedback loop ("Let each result inform the next"), an explicit running-checklist checkpoint (step 6), and error-recovery guidance (graceful degradation on tool failure); the destructive/batch cap does not apply since this is read-only investigation.	5 / 5
Progressive Disclosure	Structure and signaling are good (clear sections plus a conditional "Load companion files when relevant" block pointing one level deep to logs.md, traces.md, errors.md, sessions.md, metrics.md), but none of these referenced files exist in the bundle (no references/ directory), so navigation leads to dead ends.	3 / 5
	Total	18 / 20 Passed

Description

83%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 description that clearly states capabilities and explicit trigger conditions with concrete, natural-language phrasing. It is slightly held back by relying on a single action verb and trigger terms that overlap with general debugging skills.

Suggestions

Vary the action verbs to broaden capability coverage (e.g., "Analyzes ... correlates ... pinpoints root cause") to push specificity toward 5.

Add a few common synonyms users say ("investigate," "debug," "troubleshoot") to the trigger clause to widen natural-term coverage.

Dimension	Reasoning	Score
Specificity	Names the domain ("observability data") and enumerates concrete data types ("logs, traces, errors, sessions, and metrics") with a concrete outcome ("find root cause and actionable evidence"), but rests on essentially one action verb ("Analyzes"), so it falls just short of the comprehensive multi-action anchor 5.	4 / 5
Completeness	It explicitly answers both what ("Analyzes observability data ... to find root cause and actionable evidence") and when ("Use when the user reports a bug, an unexpected behavior, or asks about patterns across application data") with concrete trigger phrases, matching the anchor 5 example.	5 / 5
Trigger Term Quality	"Use when the user reports a bug, an unexpected behavior, or asks about patterns across application data" captures natural user phrasings, but misses common synonyms like "investigate," "debug," or "troubleshoot," leaving a few natural terms out.	4 / 5
Distinctiveness Conflict Risk	The observability-data framing carves a clear niche, but the triggers ("reports a bug," "unexpected behavior") overlap somewhat with general debugging or code-review skills, leaving minor conflict risk.	4 / 5
	Total	17 / 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: launchdarkly/ai-tooling
Path: skills/observability/investigate/SKILL.md
Commit: ef54971

Reviewed: 6 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.