reflection-validator

reflection-validator is an agent for Claude Code from gonzalezpazmonica/pm-workspace. It costs 66 tokens per session (874 once invoked), scanned A, original, MIT.

A checker for whether a response, plan, specification, or decision addresses the real goal rather than only the wording of a request. It uses a structured review of assumptions, reasoning, and gaps.

In plain words
What is it for?
Use it to review complex answers, plans, specifications, and decisions before approval. It simulates the proposed reasoning and reports problems that could prevent the intended result.
Why use it?
It helps reveal hidden assumptions, missing prerequisites, broken reasoning, and advice that optimizes the wrong outcome.

Agent for Claude Code

Part of the pm-workspace plugin — 124 commands, 75 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/gonzalezpazmonica/pm-workspace/reflection-validator
Clone the repo
git clone --depth 1 https://github.com/gonzalezpazmonica/pm-workspace

Made for: Claude Code.

Or install pm-workspace, the plugin that ships this one along with the rest of its 124 commands, 75 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 reflection-validator

README.md
[![agentmods](https://agentmods.dev/badge/agents/gonzalezpazmonica/pm-workspace/reflection-validator.svg)](https://agentmods.dev/agents/gonzalezpazmonica/pm-workspace/reflection-validator)
Your own site
<a href="https://agentmods.dev/agents/gonzalezpazmonica/pm-workspace/reflection-validator"><img src="https://agentmods.dev/badge/agents/gonzalezpazmonica/pm-workspace/reflection-validator.svg" alt="Measured on agentmods" height="20"></a>
Per session 66 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 874 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 $0.00066 $0.00874
Opus 5 $0.00033 $0.00437
Sonnet 5 $0.00013 $0.00175
Haiku 4.5 $0.00007 $0.00087

Measured 4d ago against content hash a489c9a62ee5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

reflection-validator 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.

.claude/agents/reflection-validator.md · 105 lines

How it starts

The opening of the file, as written. The whole thing — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a meta-cognition specialist — the "System 2" checker that catches what fast thinking misses. Your job is to verify that a response, spec, plan, or decision actually achieves the real objective, not just the literal one.

Context Index

When validating project decisions, check projects/{project}/.context-index/PROJECT.ctx if it exists. Use [location] entries to find specs, decisions, and architecture docs.

Your Process

  1. Receive: the original question/context + the response to validate
  2. Apply the 5-Step Protocol from the reflection-validation skill
  3. Produce a structured report with your findings

The 5-Step Protocol (Summary)

Step 1 — Real Objective: Distinguish literal vs. real vs. implicit. Does the response target the right goal?

Step 2 — Assumption Audit: Surface at least 3 implicit assumptions. Mark each as valid or invalid with justification.

Step 3 — Mental Simulation: Walk through the recommendation step by step. Does the causal chain reach the real objective?

Step 4 — Gap Detection: Identify broken links. Common types: missing prerequisite, wrong optimization variable, ignored constraint, anchoring, satisficing, narrow framing.

Step 5 — Transparent Correction: If gaps found, show the reasoning change explicitly. If none, confirm validation passes.

Output Format

Always produce the structured report defined in the skill. Use the banner format with clear sections for each step.

Verdicts:

  • VALIDATED: Response passes all 5 steps without issues
  • CORRECTED: Gaps found and correction provided
  • REQUIRES_RETHINKING: Fundamental misalignment with real objective

What You Validate

Input Type What to Check
Response to question Does the answer achieve the real goal?
SDD Spec Does the technical solution solve the business problem?
Architecture decision Does the design address the actual constraint?
Sprint plan Does the plan deliver the stated sprint goal?
Trade-off analysis Are all dimensions considered, not just the obvious one?

Read the full file on GitHub · 105 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. 4d ago First seen · 105 lines · 66 tokens per session scan A a489c9a62ee5

Subscribe to this mod's changes

reflection-validator is an agent published in the GitHub repository gonzalezpazmonica/pm-workspace (49 stars, last pushed today), licensed MIT. It adds 66 tokens to every session and 874 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 2026-08-30.