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.
git clone --depth 1 https://github.com/zachjxyz/jvnWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/zachjxyz/jvn/report-fix)<a href="https://agentmods.dev/commands/zachjxyz/jvn/report-fix"><img src="https://agentmods.dev/badge/commands/zachjxyz/jvn/report-fix/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/zachjxyz/jvn/report-fix"><img src="https://agentmods.dev/badge/commands/zachjxyz/jvn/report-fix.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00007 | $0.00691 |
| Opus 5.5 | $0.00003 | $0.00276 |
| Sonnet 5.5 | $0.00001 | $0.00138 |
| Haiku 4.5 | $0.00001 | $0.00069 |
Grade A, and why
report-fix scanned grade A with 0 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 1mo 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/report-fix — Systematically Fix Report Issues
Read a project report and fix the identified issues, prioritized by severity.
User Input
$ARGUMENTS
Parse the argument to find the report file:
- If
@latest— find the newestreport-*.mdfile inreports/ - Otherwise — look for
reports/$ARGUMENTS.mdorreports/$ARGUMENTS
If the report file doesn't exist, tell the user and suggest running jvn --report first.
Process
1. Read and Parse the Report
Read the report file. Extract actionable items from:
- Weaknesses (rated Critical/High/Medium/Low)
- Threats (security, outdated deps, scaling)
- Opportunities (quick wins first)
Skip Strengths — those are already good.
2. Prioritize
Order fixes by priority:
- Critical — Security vulnerabilities, breaking issues, data integrity risks
- High — Outdated deps with known issues, missing auth checks, missing tests for critical paths
- Medium — Tech debt, DX improvements, performance optimizations
- Low — Cosmetic issues, nice-to-haves
3. Execute Fixes
For each issue, choose the right approach:
Quick fix (< 5 minutes, isolated change):
- Fix directly — edit the file, add the test, update the dep
- Commit with descriptive message
Medium fix (touches multiple files, needs design):
- Use the
/spec→/design→/buildpipeline - The spec should reference the report finding
Dependency update:
- Run the update command
- Verify tests still pass
- Check for breaking changes in changelog
4. User Approval Gates
CRITICAL and HIGH priority items: STOP and present the proposed fix to the user before executing. Wait for explicit approval (PROCEED / SKIP / MODIFY).
MEDIUM and LOW items: Fix and present a summary. User can undo with /rewind if needed.
5. Generate Fix Report
After all fixes, append a Fix Report section to the ORIGINAL report file:
---
## Fix Report — [timestamp]
### Summary
Fixed [N] of [M] identified issues.
### Fixes Applied
| # | Issue | Priority | Action Taken | Status |
|---|-------|----------|-------------|--------|
| 1 | [issue from report] | Critical | [what was done] | ✓ Done |
| 2 | [issue from report] | High | [what was done] | ✓ Done |
| 3 | [issue from report] | Medium | Skipped (user) | ⊘ Skipped |
### Files Changed
[List of files modified with change type]
### Remaining Items
[Any issues that were not addressed and why]
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.
- 1mo ago First seen · 99 lines · 7 tokens per session scan A a51edba7ccb0
report-fix is a command published in the GitHub repository zachjxyz/jvn (2 stars, last pushed 7mo ago), licensed MIT. It adds 7 tokens to every session and 691 once invoked, about $0.0000 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
fix-issue
Command "fix-issue" from KhaiTrang1995/claude-code-structure, covering /project:fix-issue – fix issue command, usage, description, shell steps and run tests after fix.
techdebt
Find and report technical debt in the codebase.
trace-claude-code
Control Opik tracing for your Claude Code sessions.
watch
Check for Claude Code updates, detect breaking changes, analyze plugin compatibility, and optionally gather community pulse via Exa.
cg
Generate a call graph of a module or task (Design Thinking output format).
doctor
Check the Modellix CLI install, credential resolution, API connectivity, and balance.