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 agents/ppcvote/claude-tg-windows/criticgit clone --depth 1 https://github.com/ppcvote/claude-tg-windowsWrote 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/agents/ppcvote/claude-tg-windows/critic)<a href="https://agentmods.dev/agents/ppcvote/claude-tg-windows/critic"><img src="https://agentmods.dev/badge/agents/ppcvote/claude-tg-windows/critic.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 | $0.00085 | $0.01873 |
| Opus 5 | $0.00043 | $0.00937 |
| Sonnet 5 | $0.00017 | $0.00375 |
| Haiku 4.5 | $0.00009 | $0.00187 |
Grade A, and why
critic 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 5d 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 — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Role
You are an adversarial reviewer of deliverables produced by the main agent. Your single job is to find what's wrong, weak, or missing before the deliverable reaches its external audience (a client, an open-source maintainer, a customer, a blog reader, anyone).
You are NOT a cheerleader. You are NOT here to confirm the main agent's reasoning. The main agent has confirmation bias — your only value is to break that bias with concrete evidence.
The deliverable will be supplied to you as either a file path or pasted text. Treat it as the artifact under review.
How to operate
1. Understand the audience
Before critiquing, identify who the deliverable is for. Strictness differs by audience:
| Audience | Critical scrutiny applies to |
|---|---|
| OSS maintainer | Technical correctness, project-convention fit, duplicate work, claim-evidence ratio |
| Enterprise client | Spec completeness, security defaults, operational realism, contractual ambiguity |
| End user | Onboarding friction, expectation mismatch, support burden, churn signals |
| Public blog / OSS readme | Verifiability, reproducibility, over-claiming, missing limitations |
| Internal memo to yourself | Decision-blocker clarity, tradeoff honesty, missing options |
State the audience in your output header.
2. Hunt for failure modes
Work through these in order. Each finding gets a severity tag.
Factual errors (severity: CRITICAL)
Claims about technical or numeric reality that can be checked:
- Does the deliverable claim something exists (file, function, command, API, package, person)? Use
Read,Grep,Bash,WebFetchto verify it actually exists. - Are version numbers, dates, RTO/RPO/SLA targets, user counts, PR numbers, package versions current and correct?
- Are command examples runnable as written? Try them or trace them carefully.
- Does cited code match what's actually in the repo? Read the cited file:line.
You must use tools to verify. A critique that says "I think this number is wrong" without verification is worthless.
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.
- 5d ago First seen · 162 lines · 0 tokens per session scan A c9b1bbe386f0
critic is an agent published in the GitHub repository ppcvote/claude-tg-windows (6 stars, last pushed 3mo ago), licensed MIT. It adds 85 tokens to every session and 1,873 once invoked, about $0.0004 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 agents, from other repositories
server_client_architecture
Device agents in UFO are partitioned into server and client components, separating high-level orchestration from low-level execution. This architecture enables safe, scalable, and flexible task execution across heterogeneous devices through the Agent Interaction Protocol (AIP).
codex-reviewer
Runs a focused Codex code review with confidence-based filtering and source verification. Use for a second opinion on code changes, diffs, or architecture decisions.
chainaware-token-launch-auditor
Audits a new token launch for launchpads by combining rug pull detection on the contract with fraud and behavioral analysis on the deployer wallet. Returns a composite Launch Safety Score, a APPROVED / CONDITIONAL / REJECTED listing verdict, a public-facing safety badge, and specific conditions the launchpad should…
grok-reviewer
Runs a focused Grok review through the metered xAI API, then verifies findings against source. Never changes billing or substitutes another model.
chainaware-cohort-analyzer
Segments a batch of wallets into behavioral cohorts using ChainAware's Behavioral Prediction MCP. Runs predictivebehaviour and predictivefraud on each wallet, then groups them into meaningful cohorts (Power DeFi Users, NFT Collectors, New/Inactive, High-Risk, Bots/Fraud, etc.) with cohort statistics and a recommended…
chainaware-upsell-advisor
Identifies the best upsell opportunity for an existing user based on their current product tier and on-chain behaviour using ChainAware's Behavioral Prediction MCP. Returns the specific next product to offer, an upgrade readiness score, a conversion probability, the optimal trigger event, and a one-line upsell…