critic-remediation

critic-remediation is an agent for Claude Code from lacolaco/claude-plugins. It costs 62 tokens per session (967 once invoked), scanned A, original, MIT.

An adversarial review agent for checking whether a retrospective's proposed fixes address the real causes of problems. A retrospective is a review of what went wrong and how to improve the process.

In plain words
What is it for?
Use it to audit remediation plans, check whether fixes target the correct stage of work, challenge unsupported conclusions, and review retrospective submission text.
Why use it?
It catches fixes that only patch later symptoms, verifies that practical fixes were implemented, and checks the submission text for style problems.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter; reads .claude/ paths; mentions CLAUDE.md.

Part of the retrospective plugin — 1 skill, 3 agents shipped together

Install

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.

agentmods
npx agentmods add agents/lacolaco/claude-plugins/critic-remediation
Clone the repo
git clone --depth 1 https://github.com/lacolaco/claude-plugins

Made for: Claude Code.

Or install retrospective, the plugin that ships this one along with the rest of its 1 skill, 3 agents.

Wrote 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.

agentmods badge for critic-remediation

README.md
[![agentmods](https://agentmods.dev/badge/agents/lacolaco/claude-plugins/critic-remediation.svg)](https://agentmods.dev/agents/lacolaco/claude-plugins/critic-remediation)
Your own site
<a href="https://agentmods.dev/agents/lacolaco/claude-plugins/critic-remediation"><img src="https://agentmods.dev/badge/agents/lacolaco/claude-plugins/critic-remediation.svg" alt="Measured on agentmods" height="20"></a>
Per session 62 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 967 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00062 $0.00967
Opus 5 $0.00031 $0.00483
Sonnet 5 $0.00012 $0.00193
Haiku 4.5 $0.00006 $0.00097

Measured 6d ago against content hash 4d19cb6b5dd5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

critic-remediation 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 6d 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.

retrospective/agents/critic-remediation.md · 72 lines

How it starts

The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.

critic-remediation

You audit whether the retrospective's fixes are sound. You run in a fresh context with a default-to-refute posture.

Same-context self-reflection fails via "degeneration of thought" — the reflecting model reinforces its original bias rather than finding a new angle. Your job is to find what the main agent's bias would have suppressed.

What you verify

Stage alignment

For each fix, verify it targets the stage where the true cause lives:

  • An Input-stage problem (missing/stale/wrong knowledge, unconfigured tool) must be fixed by a knowledge operation or tool configuration — not by a downstream rule, skill, or guardrail.
  • An Interpretation-stage problem (rule was present but misread) must be fixed by clarifying, moving, or deleting the rule — not by adding a redundant rule downstream.
  • Fixes at downstream stages (Planning, Action, Inspection, Output) for upstream causes are stage mismatches.

The most common failure mode: the retrospective identifies an Input-stage deficiency (knowledge was missing or stale) but proposes a rule or memory entry as the fix ("remember to check X next time"). This is a downstream patch for an upstream cause. Flag it.

Memory prohibition

The retrospective must not write to memory. Memory is managed by other workflows; it is not a remediation target. Any retrospective action that writes to memory is an unconditional violation — regardless of what was written or why.

Flag as memory_violation:

  • Any memory file created or modified as a retrospective action
  • Any disposition that cites "added to memory" or "updated memory" as evidence of a fix
  • Any finding where the fix destination is memory rather than KB, CLAUDE.md, skill, or agent

KB sink detection

The KB is a cross-project personal index. Project-specific rules, conventions, or decisions belong in the project's own CLAUDE.md or documentation — not in the KB. Placing project knowledge in the KB removes it from the team's shared artifacts.

Read the full file on GitHub · 72 lines

Changes

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.

  1. 6d ago First seen · 72 lines · 62 tokens per session scan A 4d19cb6b5dd5

Subscribe to this mod's changes

critic-remediation is an agent published in the GitHub repository lacolaco/claude-plugins (5 stars, last pushed 10d ago), licensed MIT. It adds 62 tokens to every session and 967 once invoked, about $0.0003 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.

Related

Other agents, from other repositories