Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/asgarovf/locusai/debuggingnpx skills add asgarovf/locusai --skill debugginggit clone --depth 1 https://github.com/asgarovf/locusaiWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00029 | $0.01319 |
| Opus 5 | $0.00015 | $0.00660 |
| Sonnet 5 | $0.00006 | $0.00264 |
| Haiku 4.5 | $0.00003 | $0.00132 |
Grade B, and why
debugging scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Sends data to an external URLmediumData exfiltration
A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.
curl -X POST http://localhost:3000/api/users -d '{"name": "test"}' Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s http://localhost:3000/api/users/123 | jq . How it starts
The opening of the file, as written. The whole thing — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Debugging
When to use this skill
- Investigating error messages or stack traces
- Diagnosing unexpected behavior
- Fixing failing tests
- Debugging production issues
- Understanding why code doesn't work as expected
Step 1: Reproduce the problem
Before fixing anything, confirm you can reproduce it:
# Run the failing command/test
npm test -- --grep "failing test name"
pytest -xvs test_file.py::test_name
# Check error logs
tail -100 logs/error.log
docker compose logs app --tail 100
Key questions:
- What exactly is the error? (exact message, stack trace)
- When did it start? (recent commit, dependency update?)
- Where does it happen? (specific input, environment, timing?)
- How often? (every time, intermittent, under load?)
Step 2: Read the error carefully
TypeError: Cannot read properties of undefined (reading 'map')
at UserList (src/components/UserList.tsx:15:23)
at renderWithHooks (node_modules/react-dom/...)
Parse the stack trace:
- Error type:
TypeError— wrong type, likely null/undefined - Message:
Cannot read properties of undefined— something is undefined - Property:
reading 'map'— trying to call.map()on undefined - Location:
UserList.tsx:15— exact file and line - Context: React component render — data likely not loaded yet
Step 3: Narrow the scope
Binary search approach
If you don't know where the bug is, bisect:
# Git bisect to find the breaking commit
git bisect start
git bisect bad # Current commit is broken
git bisect good abc123 # This commit was working
# Git will checkout middle commits — test each one
git bisect run npm test
Isolate variables
- Does it fail with minimal input?
- Does it fail in a fresh environment?
- Does it fail without feature flags / config?
- Does it fail with the previous version of dependencies?
Step 4: Inspect state
Check the data
# Search for where the variable is set
grep -rn "userData\s*=" --include="*.ts" src/
# Check database state
psql -c "SELECT * FROM users WHERE id = 123"
# Check API response
curl -s http://localhost:3000/api/users/123 | jq .
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 189 lines · 29 tokens per session scan B 52e823d7bc6b
debugging is a skill published in the GitHub repository asgarovf/locusai (23 stars, last pushed 5mo ago), licensed MIT. It adds 29 tokens to every session and 1,319 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
seedance-troubleshoot
This skill should be used when a Seedance 2.0 output is blurry, jittery, off-prompt, morphing, blocked, visually generic, unstable, desynced, inconsistent, or otherwise fails and needs root-cause diagnosis.
huawei-cloud-mrs-host-alarm-diagnose
Huawei Cloud MRS cluster alarm diagnosis skill. Analyzes the root cause of an MRS alarm based on user-provided alarm information (alarm ID, alarm name, alarm details, occurrence time, node IP, related service and logs), then outputs the root cause, repair steps, and verification method. Diagnosis is driven by the…
errors
Handle UptimeRobot MCP errors — codes, meanings, retry policy, and user-facing recovery suggestions.
Error Log Summarizer
Parses raw error logs and produces a concise, prioritized summary of unique issues with root cause hints.
Stack Trace Analyzer
Interprets error stack traces to pinpoint root cause, explain what went wrong, and suggest fixes.
debugging
Use when debugging test failures, runtime errors, or unexpected behavior in the GAC project. Covers pytest debugging, logging, and common patterns.