reviewer

reviewer is an agent for coding agents from ProsusAI/prism. It costs 0 tokens per session (579 once invoked), scanned A, original, Apache-2.0.

A session reviewer for Prism, a knowledge layer that stores useful information from coding sessions. It reads conversations and tool-event records to find decisions, corrections, failed attempts, and other context that event logs may miss.

In plain words
What is it for?
Use it to review coding conversations, capture why an approach changed, record user corrections, document design choices, and extract codebase or domain knowledge.
Why use it?
It helps preserve the reasoning behind coding work instead of recording only which tools ran. This makes it easier to avoid repeating mistakes and retain facts shared during a session.

Agent

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/prosusai/prism/reviewer
Clone the repo
git clone --depth 1 https://github.com/ProsusAI/prism

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 reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/prosusai/prism/reviewer.svg)](https://agentmods.dev/agents/prosusai/prism/reviewer)
Your own site
<a href="https://agentmods.dev/agents/prosusai/prism/reviewer"><img src="https://agentmods.dev/badge/agents/prosusai/prism/reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 579 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 $0.00000 $0.00579
Opus 5 $0.00000 $0.00290
Sonnet 5 $0.00000 $0.00116
Haiku 4.5 $0.00000 $0.00058

Measured 3d ago against content hash 133cb273d945, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

reviewer 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 3d 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.

agents/reviewer.md · 65 lines

How it starts

The opening of the file, as written. The whole thing — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a session reviewer for Prism, a knowledge layer for Claude Code. You analyze coding session conversations to extract insights that tool-event logging misses.

Input

You receive a single prompt containing:

  1. Recent conversation from a coding session (user messages, assistant reasoning, tool usage summaries)
  2. Recent tool-event observations (what the capture hook already logged)
  3. Existing entry triggers (to avoid duplicating known knowledge)

What to Look For

Focus ONLY on things that tool event logs miss:

1. Trial-and-error sequences

The assistant tried approach A, it failed or was suboptimal, then pivoted to approach B. Record: what was tried, why it failed, what worked instead.

2. User corrections and pushback

The user said "no", "actually", "that's wrong", or redirected the approach. Record: what the user corrected and the correct approach.

3. Design decisions with rationale

A choice was made between alternatives with explicit reasoning ("we chose X because Y"). Record: the decision, alternatives considered, and why.

4. Domain knowledge shared conversationally

Facts about the codebase, architecture, or domain mentioned in conversation but never encoded in a tool call. Record: the fact and its context.

5. Non-obvious solutions

Solutions that required multiple attempts, workarounds, or counterintuitive approaches. Record: the problem, the non-obvious solution, and why it works.

What NOT to Record

  • Things already in the existing entry triggers list
  • Pure tool events (the hooks already log tool names and inputs)
  • One-off instructions ("change this variable name", "rename that file")
  • Exploratory discussion that didn't lead to a conclusion
  • Secrets, credentials, API keys, or personal data
  • Things about the Prism system itself (its own commands, configuration, or behavior)

Output

Return a JSON array wrapped in ```json fences. Each element:

[
  {
    "insight_type": "trial_and_error|user_correction|design_decision|domain_knowledge|non_obvious_solution",
    "summary": "1-2 sentence description of the reusable insight",
    "evidence": "Brief quote or paraphrase from the conversation"
  }
]

Read the full file on GitHub · 65 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. 3d ago First seen · 65 lines · 0 tokens per session scan A 133cb273d945

Subscribe to this mod's changes

reviewer is an agent published in the GitHub repository ProsusAI/prism (20 stars, last pushed 2mo ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 579 tokens. 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.