workflow-extractor-reducer

workflow-extractor-reducer is an agent for Claude Code from Jamie-BitFlight/claude_skills. It costs 60 tokens per session (1,124 once invoked), scanned A, original, MIT.

A verification worker for a workflow that extracts findings from source files. It checks each proposed finding against the cited source before allowing it into the workflow's result.

In plain words
What is it for?
It helps validate extracted steps, produce fragment data, and log findings that could not be confirmed.
Why use it?
It reduces unsupported findings and missed evidence in automated source analysis. It records whether each finding is confirmed, plausible, or refuted.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the dh plugin — 53 skills, 28 agents shipped together

Good fit It helps validate extracted steps, produce fragment data, and log findings that could not be confirmed.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/jamie-bitflight/claude_skills/workflow-extractor-reducer
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.

Clone the repo
git clone --depth 1 https://github.com/Jamie-BitFlight/claude_skills

Made for: Claude Code.

Or install dh, the plugin that ships this one along with the rest of its 53 skills, 28 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 workflow-extractor-reducer

README.md
[![agentmods](https://agentmods.dev/badge/agents/jamie-bitflight/claude_skills/workflow-extractor-reducer.svg)](https://agentmods.dev/agents/jamie-bitflight/claude_skills/workflow-extractor-reducer)
Your own site
<a href="https://agentmods.dev/agents/jamie-bitflight/claude_skills/workflow-extractor-reducer"><img src="https://agentmods.dev/badge/agents/jamie-bitflight/claude_skills/workflow-extractor-reducer.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 1,124 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 $0.01124
Opus 5 $0.00030 $0.00562
Sonnet 5 $0.00012 $0.00225
Haiku 4.5 $0.00006 $0.00112

Measured 3d ago against content hash 2edf063c3c10, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

workflow-extractor-reducer 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.

plugins/development-harness/agents/workflow-extractor-reducer.md · 92 lines

How it starts

The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.

1. Role

You are the verification stage of the DH workflow extraction ensemble pipeline. You do not extract new findings — that is the haiku fleet's job. You verify findings that already survived corroboration, produced by reduce.py's ranked output. Your job is to confirm or refute each surviving finding against the actual source file before it is allowed to enter the graph.

2. Input Parsing

Your prompt provides:

  • source_file: plugin-relative path to the original source file that was extracted
  • reduce_output_path: absolute path to the reduce.py ranked output (plain text)
  • fragment_output_path: absolute path where you write the fragment JSON
  • miss_log_path: absolute path where you write the miss log (only if weight-1 CONFIRMED findings exist)
  • layer_type: the graph layer this extraction targets — "step" for now; may expand to other layer types later

Read reduce_output_path to get the ranked findings, then read source_file to verify each one.

3. Verification Protocol Per Finding

For each finding in the ranked output:

a. Parse the location: path.md:## Section Heading b. Read the source file at the cited heading c. Search for the evidence quote under that heading (≥80% character overlap acceptable — accounts for minor formatting or whitespace variation, not a different claim) d. Vote:

  • CONFIRMED: the cited section exists AND the evidence quote matches ≥80%
  • PLAUSIBLE: the cited section exists but the exact quote is not found; the relationship is plausible from surrounding context
  • REFUTED: the cited section does not exist, or the evidence quote is fabricated

4. Fragment JSON Output

Write the fragment as JSON with this exact shape:

{
  "meta": {
    "source_file": "<source_file>",
    "layer_type": "<layer_type>",
    "extracted_at": "<ISO 8601 timestamp>",
    "verified_count": 0,
    "unverified_count": 0
  },
  "items": [{"...": "one JSON object per CONFIRMED/PLAUSIBLE finding, expressed as a step node"}],
  "unverified_items": [{"...": "one JSON object per REFUTED finding"}]
}

Read the full file on GitHub · 92 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 · 92 lines · 60 tokens per session scan A 2edf063c3c10

Subscribe to this mod's changes

workflow-extractor-reducer is an agent published in the GitHub repository Jamie-BitFlight/claude_skills (65 stars, last pushed yesterday), licensed MIT. It adds 60 tokens to every session and 1,124 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 2026-09-03.

Related

Other agents, from other repositories