/cco-coach
Analyze the user's last prompt for clarity, scope and specificity — give a quality score and concrete suggestions to make the next prompt produce better results
$ npx -y skills add egorfedorov/claude-context-optimizer --skill cco-coach --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/cco-coach
Context preview
The summary Claude sees to decide when to auto-load this skill.
Analyze the user's last prompt for clarity, scope and specificity — give a quality score and concrete suggestions to make the next prompt produce better results
SKILL.md
cco-coach.SKILL.mdname: cco-coach
description: Analyze the user's last prompt for clarity, scope and specificity — give a quality score and concrete suggestions to make the next prompt produce better results
license: MIT
allowed-tools: [Bash, Read]
Prompt Coach
When the user runs `/cco-coach` or asks for "prompt feedback" / "оцени мой промпт", run the analyzer on the user's last natural-language ask (or on text they provide).
How to use
If the user passes text after the command (e.g. `/cco-coach add login screen`):
node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js analyze "the prompt text here"If no text given, show recent prompt history and average score:
node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js history 10What the score means
The coach grades on four dimensions (S/A/B/C/D/F):
- **Specificity** — does the prompt name files, line numbers, identifiers?
- **Scope** — is the change bounded? "fix bug in src/auth.js:42" beats "fix all bugs"
- **Success criteria** — how do we know it's done? (tests pass, error gone, etc.)
- **Length** — too short usually means under-specified; too long buries the ask
Presentation
After running the script, summarise the top 1–3 suggestions in plain language and offer to **rewrite the prompt** for the user. If they say yes:
node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js improve "the original prompt"Then show the improved version with a brief diff explaining what changed.
When to suggest this proactively
If you notice the user's prompt scored ≤60 in the last UserPromptSubmit hook and a task is going off the rails (lots of reads with no edits), gently suggest `/cco-coach` to refine.
Read more
name: cco-coach description: Analyze the user's last prompt for clarity, scope and specificity — give a quality score and concrete suggestions to make the next prompt produce better results license: MIT allowed-tools: [Bash, Read]
Prompt Coach
When the user runs `/cco-coach` or asks for "prompt feedback" / "оцени мой промпт", run the analyzer on the user's last natural-language ask (or on text they provide).
How to use
If the user passes text after the command (e.g. `/cco-coach add login screen`):
node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js analyze "the prompt text here"If no text given, show recent prompt history and average score:
node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js history 10What the score means
The coach grades on four dimensions (S/A/B/C/D/F):
- **Specificity** — does the prompt name files, line numbers, identifiers?
- **Scope** — is the change bounded? "fix bug in src/auth.js:42" beats "fix all bugs"
- **Success criteria** — how do we know it's done? (tests pass, error gone, etc.)
- **Length** — too short usually means under-specified; too long buries the ask
Presentation
After running the script, summarise the top 1–3 suggestions in plain language and offer to **rewrite the prompt** for the user. If they say yes:
node ${CLAUDE_PLUGIN_ROOT}/src/prompt-coach.js improve "the original prompt"Then show the improved version with a brief diff explaining what changed.
When to suggest this proactively
If you notice the user's prompt scored ≤60 in the last UserPromptSubmit hook and a task is going off the rails (lots of reads with no edits), gently suggest `/cco-coach` to refine.
Claude Code plugin that tracks token usage, identifies wasted context, and saves 30-50% on API costs. Heatmaps, ROI reports, budget alerts, efficiency scores, git-aware suggestions — all local, zero config.
Repo: egorfedorov/claude-context-optimizer
Other skills on claude-context-optimizer.
- /cco-anatomy
Generate a compact project map so Claude understands the codebase without opening every file
Open skill - /cco-budget
Configure token budget limits, auto-compact settings, and view current budget status (model-aware — Opus 4.8 default, full 1M context at standard price)
Open skill - /cco-claudemd
Analyze CLAUDE.md files for token bloat and suggest optimizations
Open skill - /cco-clean
Clean up old tracking data and reset statistics
Open skill - /cco-digest
Show weekly/daily efficiency digest with score and trends
Open skill - /cco-doctor
Health check for the context-optimizer plugin install — verifies versions, hooks, data dir, model config, and reports any issues
Open skill

