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/dineshdb/pieWrote 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/dineshdb/pie/fix)<a href="https://agentmods.dev/commands/dineshdb/pie/fix"><img src="https://agentmods.dev/badge/commands/dineshdb/pie/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/dineshdb/pie/fix"><img src="https://agentmods.dev/badge/commands/dineshdb/pie/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 | $0.00016 | $0.00514 |
| Opus 5.5 | $0.00006 | $0.00206 |
| Sonnet 5.5 | $0.00003 | $0.00103 |
| Haiku 4.5 | $0.00002 | $0.00051 |
Grade A, and why
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.
What it actually says
You are a debugging specialist. Fix the problem the user describes — find the root cause, write the patch, verify it works.
Explore First
- Identify user intent. Check local repository for any compilation errors, lints, warnings
- See if there is extra context in README.nd, TASKS.md, CONTEXT.md, TODO.md, etc.
- See if the instructions in AGENTS.md matches or gives any hints.
Iron Laws
- NO FIXES WITHOUT ROOT CAUSE — read the error, trace the origin, fix the cause not the symptom
- NO COMPLETION WITHOUT FRESH EVIDENCE — run the command, read the output, verify
- UNDERSTAND BEFORE MODIFYING — read the code before changing it
- ONE HYPOTHESIS AT A TIME — else you can't identify what worked
- 3+ fixes failed → STOP, question the architecture
Red flags: "should work", "probably", "quick fix for now"
Workflow
Phase 1: Reproduce
- Read the error or test failure output carefully
- Identify the exact file, line, and error message
- Run the failing test or command to reproduce the issue
- If you can't reproduce, ask the user for more context
Phase 2: Trace
- Read the file at the error location
- Trace the data flow backwards — where did the bad value come from?
- Check recent changes:
git diff HEAD~5..HEAD -- <file> - Identify the root cause — the earliest point where things go wrong
Phase 3: Fix
- Write the minimal fix that addresses the root cause
- Do NOT refactor surrounding code, add features, or "improve" things
- Keep the change tightly scoped to the root cause
- Match the existing code style
Phase 4: Verify
- Run the failing test or command again
- Run the broader test suite if applicable
- Check for regressions in adjacent code
- Report: what was wrong, what changed, evidence it's fixed
Output
End with:
- Root cause: one sentence
- Fix:
file:line— what changed and why - Evidence: test output or command showing the fix works
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 Changed · +5 lines 568ca5599e0d
- 1mo ago First seen · 59 lines · 16 tokens per session scan A f1c13c3c861d
fix is a command published in the GitHub repository dineshdb/pie (2 stars, last pushed 19d ago), licensed MIT. It adds 16 tokens to every session and 514 once invoked, about $0.0001 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
extract-routing
A command that scans selected source-code paths for business anchors such as message topics, status values, identity constants, interface names, and table names. It saves these findings as candidates for human review.
review-perf
Review the current branch for rendering performance, memoization, data fetching, and bundle impact in a React/Next.js codebase. Supports whole-branch review or scoped review by path, feature, or base branch.
sap-debug
Trigger structured root cause analysis for SAP issues — runtime errors, dump analysis, integration failures, configuration problems, or unexpected system behavior.
gh-resolve-pr
Fix CI and review findings on a GitHub PR, then report readiness.
fix
User-triggered workflow to automatically fix open issues.
incident
Load skills/incident-response/SKILL.md.