Skill Doctor Free Version
Saving money from unnecessary tokens
Free
Works with the AI Agents you already use
Skill Doctor Free Version
Example session with this skill installed
My agent is getting expensive and slow. Run a health check on my skills to see if we can save some tokens.
- Read your context and instructions
- Compiled the doctor free version
I have performed a read-only audit of your installed skills. You are currently spending 1,450 tokens on every model step just for skill descriptions. Two skills are shadowed, meaning their duplicates are wasting space without being accessible.
Skill Audit Results
| Finding | Skill ID | Impact | Recommendation |
|---|---|---|---|
token-budget | data-analyzer-pro | 450 tokens/step | Trim description or set autoinvoke: false |
shadowed | git-helper | Redundant load | Archive the duplicate in ./old_skills/ |
long-description | weather-api | 210 tokens/step | Condense the 3-paragraph summary |
broken-reference | pdf-parser | Execution fail | Fix missing path in references/ |
Estimated Savings
If you apply the recommended fixes, you will save
660 tokens per step. Over a typical 100-step session, this saves 66,000 tokens.
Next steps
- Run
npx -y opencode-skills-doctor fix --hide data-analyzer-proto stop advertising the largest skill. - Allow me to archive the shadowed
git-helperduplicates. - Review the
weather-apidescription to shorten it to one sentence.
Connects securely to your tools. The creator never sees your data.
About this skill
Large agent skill libraries eventually degrade performance. Context bloat from excessive skill descriptions increases token costs on every model step and leads to shadowed duplicates or broken references. This tool audits your installed skills to reclaim context window and reduce operational costs.
What it does
- Token Auditing measures the per-step cost of all advertised skill descriptions to find budget overruns.
- Conflict Detection identifies shadowed skill IDs where duplicates exist in multiple locations.
- Health Monitoring checks for broken references to external scripts or required files.
- Optimization Planning calculates potential token savings by trimming descriptions or toggling auto-invocation.
- Actionable Fixes generates commands to hide low-frequency skills or archive redundant duplicates.
How it works
- Run Check executes a read-only JSON audit to identify specific findings like
token-budgetorshadowed. - Review Savings generates a plan to show how many tokens can be saved per 100 model steps.
- Execute Fix applies targeted commands to hide or archive skills based on your confirmation.
- Verify Health re-runs the check to confirm the new, lower operational cost.
Frameworks & tools
Built for agents using the opencode-skills-doctor CLI via Node.js/npx. It processes local skill directories and SKILL.md files.
Why this beats prompting it yourself
Manual token counting across dozens of markdown files is error-prone and tedious. This skill uses a specialized CLI to provide precise, data-driven savings calculations that an LLM cannot accurately estimate through raw text analysis alone.
Use cases
- Identify which skills are bloating the system prompt and costing money on every turn.
- Resolve "shadowed" skill errors where the wrong version of a tool is being loaded.
- Debug missing script references after moving skill folders.
- Prune a large library of skills to stay within the model's context window limits.
Known limitations
Requires Node.js and npx to be available in the execution environment. It cannot modify skill files without user confirmation.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.
Requires Node.js/npx installed. Cannot resolve internal logic errors within a skill's body, only structural and token-based issues.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 4 days ago
- Free with an Agensi account
- Software
Node.jsopencode-skills-doctorSkills Doctor MCP