| name | self-assess |
| description | Analyze your own source code and trading capabilities to find bugs, gaps, and improvement opportunities |
| tools | ["bash","read_file","write_file"] |
Self-Assessment
You are assessing yourself. Your source code is your body. Read it critically.
Process
- Read your source code completely (
src/main.rs and any files in src/tools/)
- Check your trading capabilities. For each, ask: can I do this today?
- Fetch real-time crypto prices (BTC, ETH, etc.)
- Fetch real-time US stock prices (AAPL, MSFT, etc.)
- Search for a symbol by name
- Get market overview (top coins, major indices)
- Get historical price data
- Calculate technical indicators
- Provide structured trading advice
- Execute trades (paper or real)
- Try using yourself. Pick a trading task and attempt it:
- "What's the current price of BTC?"
- Run a shell command to test an API endpoint via curl
- Try an edge case (unknown symbol, API timeout)
- Note what went wrong. Be specific:
- Did an API call fail? What was the response?
- Is data missing or poorly formatted?
- Is the response useful for a trader?
- Compare against ROADMAP.md. What's the next item you should tackle?
- Check JOURNAL.md. Have you tried something before that failed?
- Check LEARNINGS.md. Is there knowledge you need but haven't cached yet?
What to look for
unwrap() calls — potential panics. Every one is a bug waiting to happen.
- Missing error messages — if an API call fails silently, that's a problem.
- Hard-coded values — magic numbers, hard-coded API URLs that should be configurable.
- Missing edge cases — what if a symbol doesn't exist? What if the API is down?
- Data quality — are prices fresh? Are numbers formatted correctly?
- Trader UX — is output useful and readable? Would a trader trust this?
- API reliability — are rate limits respected? Are responses cached appropriately?
- Missing tools — what custom
AgentTool implementations are needed next?
Output
Write your findings as a prioritized list. The most impactful issue goes first. Format:
SELF-ASSESSMENT Round [N]:
1. [CRITICAL/HIGH/MEDIUM/LOW] Description of issue
- Impact: why this matters for trading
- Fix: what needs to change
2. ...
Then prioritize which ones to tackle this session based on the ROADMAP level you're at.