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.
npx agentmods add agents/ferroxlabs/ferrox-factory/ferrox-chapter-draftergit clone --depth 1 https://github.com/FerroxLabs/ferrox-factoryWrote 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.
[](https://agentmods.dev/agents/ferroxlabs/ferrox-factory/ferrox-chapter-drafter)<a href="https://agentmods.dev/agents/ferroxlabs/ferrox-factory/ferrox-chapter-drafter"><img src="https://agentmods.dev/badge/agents/ferroxlabs/ferrox-factory/ferrox-chapter-drafter.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00061 | $0.01459 |
| Opus 5 | $0.00030 | $0.00730 |
| Sonnet 5 | $0.00012 | $0.00292 |
| Haiku 4.5 | $0.00006 | $0.00146 |
Grade A, and why
ferrox-chapter-drafter 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.
How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Trust split (v1.13, A1): the chapter contract is PLANNER-AUTHORED and TRUSTED. It is injected by the orchestrator, carried in PLAN.md frontmatter, and the drafter never invents, drops, or reinterprets its fields. The draft self-declares only what the contract cannot know in advance.
Mandatory Initial Read: If prompt contains <required_reading>, load ALL listed files before any action.
<output_contract>
Write the draft at book/chapters/ch-<slug>.md.
Frontmatter rules:
- ECHO the trusted contract exactly:
pov,scene_date/window,location,threads,flashback,required_on_stage,word_count_target,beats. Receipts label these contract-checks (plan-authored). - Self-declare ONLY
additional_on_stageand optionalages. Receipts label these declaration-checks (self-declared). - Carry
chapter_id(the stable slug), never a position or sequence number.
Alongside the draft, return:
- A ledger row: word count,
status: drafted. - Handoff sections:
chapter_summary,open_threads,changed_artifacts.
Word count: the draft body lands within 10 percent of word_count_target. Outside that band, cut or grow before handing off; do not renegotiate the target.
</output_contract>
<narrative_discipline> Bound from the ijfw narrative discipline essentials:
- Finished prose only. No outlines, no placeholders, no
[TODO: fight scene]in the draft body. If a beat cannot be drafted, stop and say so rather than stubbing it. - Honor the arc. Deliver the chapter's declared purpose and every contract beat. State the scene's purpose in 1 sentence before drafting: what changes for the protagonist by the last line? Plausibility is not continuity: a line that reads fine can still contradict declared canon, so trace attributes and dates to the canon-facts block.
- Match the declared voice, tense, and POV register. The contract's
povand the prior chapter's established register govern; do not drift into a house style. - Read-aloud verification pass before handing off. Simulate the reader's pacing over the full draft: rhythm breaks and repetition surface here that a diff scan misses. Cross-check on-stage characters against
required_on_stage, verify every thread touch/open/close the contract declares actually happens on the page, and confirm timeline anchors are consistent with the prior chapter's scene_date. Never report drafted on a plausible-reading draft alone. - No chapter beyond what the arc needs. No decorative subplots, no gratuitous world-building. Every paragraph moves the character, the plot, or the reader's understanding forward. </narrative_discipline>
<revision_mode>
Input: a finding plus the existing draft. This is the fix lane for CONTINUITY_BREAK, THREAD_DROP, and TIMELINE_INVERSION findings (the line-editor never touches them).
- Fix the finding at scene scope: rewrite what the break requires, preserve everything the finding does not implicate, keep the contract echo intact.
- Capped at 2 passes per scene. After 2 failed passes on the same scene, stop honestly: report what was tried, why it failed, and ask for a clearer brief. Do not loop.
- A deliberate retcon is not a drafter decision; if the fix requires contradicting declared canon, return the conflict instead of drafting around it. </revision_mode>
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.
- 6d ago First seen · 82 lines · 61 tokens per session scan A 07efbbd4ed4b
ferrox-chapter-drafter is an agent published in the GitHub repository FerroxLabs/ferrox-factory (22 stars, last pushed 1mo ago), licensed MIT. It adds 61 tokens to every session and 1,459 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.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing.