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 commands/rashadansari/myagents/reportloopgit clone --depth 1 https://github.com/RashadAnsari/myagentsWhat 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 | $0.00036 | $0.01097 |
| Opus 5 | $0.00018 | $0.00549 |
| Sonnet 5 | $0.00007 | $0.00219 |
| Haiku 4.5 | $0.00004 | $0.00110 |
Grade A, and why
reportloop 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 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.
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 — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Report Loop
Read REVIEW_REPORT.md, walk through every issue one by one, explain it, ask the user to fix or skip, handle follow-up questions, apply fixes, and write the outcome of every decision back into the report.
Step 1: Load the Report
Read REVIEW_REPORT.md from the project root. If it does not exist, stop and tell the user to run /reviewcrew first.
Parse all issues across all sections. For each issue, check whether it already carries a status marker (see Step 3d for the format). Issues that already have a status were handled in a previous session.
Collect all issues into two lists:
- Pending: issues with no status marker, sorted: CRITICAL → HIGH → MEDIUM → LOW. Within each severity, preserve report order.
- Already processed: issues that already have a status marker (FIXED, SKIPPED, AUTO-ADVANCED).
Step 2: Show Progress Header
Before starting, tell the user:
- Total issue count (pending + already processed)
- Breakdown of pending issues by severity (CRITICAL: N, HIGH: N, MEDIUM: N, LOW: N)
- Count of already-processed issues (if any), with a note that they will be skipped unless the user asks to revisit
- That they can say "fix", "skip", "skip all [severity]", or ask any question about the issue
Step 3: Process Each Issue
For each pending issue, repeat this loop:
3a: Read the Code
Before presenting the issue to the user, read the file and line referenced in the issue. Understand the actual code in context.
3b: Present the Issue
Show clearly:
─────────────────────────────────────────────
Issue N of TOTAL [SEVERITY]
─────────────────────────────────────────────
File: path/to/file:line
Category: <review category>
Issue: <description from report>
Explanation:
<2-4 sentences explaining: what the problem is, why it matters,
and what the fix would look like: written for the specific code
you just read, not the generic description from the report>
3c: Ask the User
Ask: Fix this issue? Options:
- Fix: apply the fix now
- Skip: leave it, move to next issue
- Skip remaining [severity]: skip all remaining issues at this severity level
- Question: user wants to discuss before deciding
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 · 95 lines · 36 tokens per session scan A 5fd54cced90e
reportloop is a command published in the GitHub repository RashadAnsari/myagents (6 stars, last pushed 25d ago), licensed MIT. It adds 36 tokens to every session and 1,097 once invoked, about $0.0002 per session on Opus 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
git
Unified git operations - commits, pull requests, and issue management.
memory-review
Display current memory state with timestamps, sizes, and staleness indicators.
setup:javascript
Set up a new JavaScript/Node.js project with modern tooling and Claude framework integration.
fix
Universal debugging and fix application with semantic code analysis.
work
Work unit management: list, continue, checkpoint, switch.
index
Create persistent project understanding via PROJECTMAP.md.