CtrlK
BlogDocsLog inGet started
Tessl Logo

agenttrace-session-audit

Audit local AI coding-agent sessions with agenttrace for cost, tokens, tool failures, latency, anomalies, health, diffs, and CI gates.

66

Quality

80%

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 ./plugins/all-skills/skills/agenttrace-session-audit/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

93%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.

The body is an exemplary lean operational skill: fully executable commands, a clear discovery-first workflow with a diagnostic checkpoint, and tight reporting guardrails. The only meaningful gap is that a few failure paths (missing binary outside the repo, empty session stores) have direction but no explicit recovery loop.

Suggestions

Add an explicit fallback step for when the agenttrace binary is missing and the working directory is not the agenttrace repository (e.g., where to install it or how to proceed).

Turn the "no sessions detected" note into a short validate-and-retry loop: run --doctor, report detected directories, then stop or continue based on findings.

DimensionReasoningScore

Conciseness

The ~57-line body is pure directive content — commands, thresholds, and reporting rules — with zero concept explanations or padding; every token earns its place, matching anchor 5. It is clearly not anchor 4, which requires some over-explanation to trim, and none is present.

5 / 5

Actionability

Every step is a copy-paste-ready command ("agenttrace --overview --fail-under-health 80 --fail-on-critical --max-tool-fail-rate 15") including exact CI-gate flags and output paths, matching anchor 5's fully executable coverage of common cases. Nothing is pseudocode or hand-wavy.

5 / 5

Workflow Clarity

The numbered Workflow section sequences discovery, binary fallback ("cargo run -q -p agenttrace --"), output formats, and per-session queries, with a "--doctor" diagnostic checkpoint — matching anchor 4's clear sequence with most checkpoints. It falls short of anchor 5 because error-recovery paths (binary missing and not in the repo, no sessions detected beyond the doctor callout) are implicit rather than explicit feedback loops.

4 / 5

Progressive Disclosure

The skill is under 50 lines with no bundle files and no external references; the Workflow / Report Focus / Guardrails sections are well-organized and self-contained, which per the simple-skill scoring note earns a 5. There is nothing buried or inlined that belongs in a separate file.

5 / 5

Total

19

/

20

Passed

Description

66%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 is concrete and specific with strong coverage of audit targets and a clear niche, but it omits an explicit "Use when..." trigger clause, which both caps completeness and weakens its usefulness as a trigger surface. Adding a when-to-use clause with natural user phrasing would lift it to the top level.

Suggestions

Append a trigger clause such as "Use when the user asks about session cost, token usage, tool failures, retry loops, or agent-trace health and CI gates" to fix the completeness gap.

Add natural synonyms users are likely to say — e.g. "spend", "usage", "token burn", "session logs" — to improve trigger term coverage.

Consider naming the actionable outputs (reports, health gates) alongside the audited metrics to sharpen the what-it-does statement.

DimensionReasoningScore

Specificity

"Audit local AI coding-agent sessions with agenttrace for cost, tokens, tool failures, latency, anomalies, health, diffs, and CI gates" enumerates eight concrete audit targets, matching anchor 4's several specific actions. It stops short of anchor 5 because it uses a single action verb (audit) rather than multiple distinct concrete actions.

4 / 5

Completeness

The "what" is clear and concrete (audit sessions for the listed metrics), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric guidelines. It scores above 2 because the "what" is fully specified rather than vague.

3 / 5

Trigger Term Quality

Includes good natural terms users would say — "cost", "tokens", "tool failures", "latency", "health", "CI gates" — plus the tool name "agenttrace". Missing common synonyms like "spend", "usage", or "token burn", so it does not reach anchor 5's comprehensive synonym coverage.

4 / 5

Distinctiveness Conflict Risk

The tool-named, niche scope ("local AI coding-agent sessions with agenttrace") makes it mostly distinct with only minor overlap risk against general observability or logging skills. Not anchor 5 because terms like "cost", "tokens", and "CI" alone could weakly match adjacent skills.

4 / 5

Total

15

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
davepoon/buildwithclaude
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.