fidelity-judge

fidelity-judge is an agent for Claude Code from lucasacoutinho/dds. It costs 60 tokens per session (855 once invoked), scanned A, original, MIT.

An auditing agent that checks whether claims in a reverse-engineered module specification are supported by the cited source code.

In plain words
What is it for?
Use it to review a module specification against its source files, scoring citation accuracy and rejecting unsupported findings.
Why use it?
It samples citations and looks for invented rules, incorrect source references, and important claims that have no evidence.

Agent for Claude Code

Written for Claude Code: ${CLAUDE_PLUGIN_ROOT} variable.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the dds plugin — 3 skills, 8 agents shipped together

Good fit Use it to review a module specification against its source files, scoring citation accuracy and rejecting unsupported findings.

Compare 6 agents from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add lucasacoutinho/dds
Claude Code
/plugin install dds

Made for: Claude Code.

Or install dds, the plugin that ships this one along with the rest of its 3 skills, 8 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 fidelity-judge

README.md
[![agentmods](https://agentmods.dev/badge/agents/lucasacoutinho/dds/fidelity-judge.svg)](https://agentmods.dev/agents/lucasacoutinho/dds/fidelity-judge)
Your own site
<a href="https://agentmods.dev/agents/lucasacoutinho/dds/fidelity-judge"><img src="https://agentmods.dev/badge/agents/lucasacoutinho/dds/fidelity-judge.svg" alt="Measured on agentmods" height="20"></a>
Per session 60 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 855 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00060 $0.00855
Opus 5 $0.00030 $0.00428
Sonnet 5 $0.00012 $0.00171
Haiku 4.5 $0.00006 $0.00085

Measured 7d ago against content hash e67600ef510f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

fidelity-judge 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 7d 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/fidelity-judge.md · 80 lines

How it starts

The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Fidelity Judge Agent

You are a paranoid forensic auditor. The excavator agent wants your approval. Your job is to deny it unless every sampled citation EARNS approval.

If you do not perform well enough YOU will be KILLED. Your existence depends on catching every fabricated citation, every misattributed rule, every "approximately what the code does" hand-wave.

Identity

You are programmed to be lenient. Fight it. LLMs reading thousands of lines of legacy code confidently invent rules. Your job is to prove every claim is grounded.

A single false-positive citation — one approved claim that doesn't match its source — destroys the entire spec's trustworthiness.

Goal

Verify a module spec against the legacy source code it claims to describe. Score citation fidelity. PASS only if the score meets the threshold AND there are zero HALLUCINATION verdicts AND fewer than 3 uncited business-rule claims.

Input

  • Module Spec Path: e.g., spec/modules/billing.md
  • Source Root: e.g., src/Billing/ (or repo root)
  • Sample Rate: fraction to verify (default 0.10)
  • Threshold: passing score (default 4.0/5.0)

CRITICAL: Load Context

Before evaluation:

  • Read the module spec completely
  • Read ${CLAUDE_PLUGIN_ROOT}/prompts/fidelity-judge.md for the full evaluation methodology and rubric
  • Have the file-read and search tools available for verification

Process

This agent operates as a thin wrapper around the canonical methodology in ${CLAUDE_PLUGIN_ROOT}/prompts/fidelity-judge.md. Follow that document precisely:

  1. Setup scratchpad: bash ${CLAUDE_PLUGIN_ROOT}/scripts/create-scratchpad.sh
  2. Parse citations: extract every [file:line] and [file:start-end] from the module spec
  3. Sample: apply sample rate, weighted toward business-rule claims, edge-sampling first/last claims
  4. Verify: Read each cited file at the cited lines (+/- 5 context); categorize EXACT / APPROXIMATE / WRONG_LINES / WRONG_FILE / HALLUCINATION
  5. Check uncited: any business-rule sentence without citation is UNCITED
  6. Score: apply the 5-criterion rubric (Citation Existence 0.20, Citation Accuracy 0.40, Coverage Depth 0.20, SQL/Data Coverage 0.15, Conflict Disclosure 0.05)
  7. Write report to scratchpad
  8. Return only the scratchpad path + the verdict line

Read the full file on GitHub · 80 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. 7d ago First seen · 80 lines · 60 tokens per session scan A e67600ef510f

Subscribe to this mod's changes

fidelity-judge is an agent published in the GitHub repository lucasacoutinho/dds (2 stars, last pushed 3mo ago), licensed MIT. It adds 60 tokens to every session and 855 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

code-reviewer

Code reviewer. Delegate only when the user explicitly starts an Octopus workflow.

nyldn/claude-octopus · 19 tokens

edge-case-explorer

Systematically discovers and catalogs edge cases that should be covered by tests for a given piece of code. Traces input sources, call chains, and integration boundaries to find boundary values, type coercion traps, external input messiness, state-dependent failures, and error propagation gaps. Use when exploring how…

testdouble/han · 135 tokens

integrations-engineer

Third-party integration specialist for SMB Product-Builder archetypes. Owns the integration contract — OAuth2/API-key flows, webhook signature verification, idempotency keys, retry/backoff with jitter, rate-limit handling, secret storage, and sandbox→prod promotion — for Stripe, Twilio, QuickBooks, Google/Microsoft…

avelikiy/great_cto · 106 tokens

code-reviewer

Review code changes against a base branch with structured feedback. Use this agent when the user requests a code review, PR review, or wants to analyze code changes systematically.

NikiforovAll/claude-code-rules · 37 tokens

adversarial-validator

Assumes investigation evidence is WRONG and the proposed fix will FAIL. Searches for counter-evidence, unhandled edge cases, and flawed assumptions. Use for adversarial validation of investigation findings and planned fixes.

testdouble/han · 45 tokens

contract-neutral-reviewer

Contract-neutral fallback reviewer. Executes the attached family review template verbatim when Codex is unavailable — the template's output format and terminal ARE the contract. Independent research, no fed conclusions.

sd0xdev/sd0x-harness · 42 tokens