caveman-stats

Show real token usage and estimated savings for the current session. Reads directly from the Claude Code session log — no AI estimation. Triggers on /caveman-stats. Output is injected by the mode-tracker hook; the model itself does not compute the numbers.

Quality

56%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Fix and improve this skill with Tessl

tessl review fix ./plugins/caveman/skills/caveman-stats/SKILL.md

Quality

Content

27%

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

This skill body is essentially a brief explanation that the model doesn't need to act, delegating everything to external hooks. While the intent (a passive/no-op skill) is understandable, the content provides no actionable guidance, no example of the stats output format, and no verifiable references to the hook files it mentions. It reads more like a comment than a skill document.

Suggestions

Add an example of the actual stats output that the hook produces, so the skill documents what the user will see and Claude can reference it if asked about the format.

Include the hook file(s) in the bundle or inline the key logic so the references to `hooks/caveman-stats.js` and `hooks/caveman-mode-tracker.js` are verifiable.

Add a minimal structure with at least a heading and a brief section explaining what triggers the skill (`/caveman-stats`) and what Claude should do if the hook fails or the user asks follow-up questions about the stats.

If Claude truly has zero responsibility, state explicitly what Claude should say if a user asks about token stats outside of the hook context (e.g., 'Refer the user to /caveman-stats').

Dimension	Reasoning	Score
Conciseness	The content is brief (3 sentences), but it's essentially all explanation of how the system works rather than actionable instruction. The phrase 'The model does not need to do anything when this skill fires' is useful context, but the rest is somewhat redundant given the description already covers this.	2 / 3
Actionability	There is no actionable guidance whatsoever — no code, no commands, no examples of output, no instructions for Claude to follow. The content merely describes what happens externally without giving Claude anything concrete to do or respond with.	1 / 3
Workflow Clarity	The workflow is implicitly described (hook fires, blocks, returns stats), but there are no explicit steps, no validation, and no detail on what the output looks like or how errors are handled. For a simple 'do nothing' skill this is borderline acceptable, but the lack of any example output or trigger details is a gap.	2 / 3
Progressive Disclosure	The content references `hooks/caveman-stats.js` and `hooks/caveman-mode-tracker.js` but no bundle files are provided, so these references are unverifiable. There's no structure, no sections, and no navigation — just a single paragraph.	1 / 3
	Total	6 / 12 Passed

Description

85%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

This is a well-crafted description that clearly defines what the skill does (show token usage and savings from session logs), when to use it (on /caveman-stats), and how it works (injected by mode-tracker hook, not AI-computed). Its main weakness is limited natural language trigger terms — users asking about token costs or usage in plain language might not trigger this skill unless they know the specific command.

Suggestions

Add natural language trigger terms like 'token cost', 'how many tokens', 'session cost', 'usage statistics' to improve discoverability beyond the slash command.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete actions: 'Show real token usage and estimated savings', 'Reads directly from the Claude Code session log', 'Output is injected by the mode-tracker hook'. Also clarifies what the model does NOT do ('no AI estimation', 'the model itself does not compute the numbers').	3 / 3
Completeness	Clearly answers both 'what' (show real token usage and estimated savings by reading session logs) and 'when' (triggers on /caveman-stats). The trigger condition is explicit and unambiguous.	3 / 3
Trigger Term Quality	Includes the explicit trigger '/caveman-stats' and relevant terms like 'token usage', 'savings', 'session'. However, it misses natural user phrases like 'how many tokens', 'cost', 'usage stats', 'how much have I spent' that users might naturally say.	2 / 3
Distinctiveness Conflict Risk	Highly distinctive with a unique slash command trigger '/caveman-stats' and a very specific niche (token usage from session logs via mode-tracker hook). Unlikely to conflict with any other skill.	3 / 3
	Total	11 / 12 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 — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: JuliusBrussee/caveman
Commit: 25d22f8

Reviewed: 20 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.