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/zach-source/mcp-rust-proxy/mcp-proxy-give-feedbackgit clone --depth 1 https://github.com/zach-source/mcp-rust-proxyWrote 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/zach-source/mcp-rust-proxy/mcp-proxy-give-feedback)<a href="https://agentmods.dev/commands/zach-source/mcp-rust-proxy/mcp-proxy-give-feedback"><img src="https://agentmods.dev/badge/commands/zach-source/mcp-rust-proxy/mcp-proxy-give-feedback.svg" alt="Measured on agentmods" 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 | $0.00008 | $0.00186 |
| Opus 5 | $0.00004 | $0.00093 |
| Sonnet 5 | $0.00002 | $0.00037 |
| Haiku 4.5 | $0.00001 | $0.00019 |
Grade A, and why
mcp-proxy-give-feedback 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 4d 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
Submit feedback on your most recent response to improve context quality.
Arguments:
- $1: Score from -1.0 (poor) to 1.0 (excellent)
- $2: Feedback comment (optional)
Example: /mcp-proxy:give-feedback 0.8 "Accurate and helpful response"
Instructions for Claude:
- Get the last tracked response_id from /tmp/mcp-proxy-last-response-id
- Use the mcp__proxy__tracing__submit_feedback tool with:
- response_id: (from file)
- score: $1
- feedback_text: $2 (if provided)
- user_id: "claude-assistant"
- Display the propagation results
- Explain which contexts were updated and by how much
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.
- 4d ago First seen · 24 lines · 8 tokens per session scan A 647540f0f637
mcp-proxy-give-feedback is a command published in the GitHub repository zach-source/mcp-rust-proxy (5 stars, last pushed 10mo ago), licensed MIT. It adds 8 tokens to every session and 186 once invoked, about $0.0000 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
VibeGuard: Learn
Dual-mode learning: (A) Extract guard rules from errors (B) Extract reusable skills from discoveries. Automatic mode routing.
avatar-orchestrator
Use the preference model to triage several blockers/decisions when is unreachable. It may execute only work that is already authorized, reversible, local, and low-impact. Read the avatar files in and follow START.md.
be-my-avatar
Run the preference model in action mode only inside authority the user already granted. Read the avatar files in and follow START.md. The model is not an authority grant.
decide-like-me
Run the preference model in decision-support mode. Read the avatar files in and follow START.md. The result is a fallible hypothesis, not a fact about the user or an authority grant.
read-my-mind
Run the user's ToM model in prediction mode (feedback precognition). Read the avatar files in and follow START.md.
report
Report a bug, request a feature, or ask a question about a heurema product. Auto-detects repo, collects context, submits GitHub issue. Triggers: "/report", "report a bug", "file an issue", "report issue".