section-writer

A writer for self-contained implementation sections in a larger development plan. It produces raw Markdown that gives an implementer enough detail to build the section without the rest of the plan.

In plain words
What is it for?
Use it to write implementation-ready plan sections with specific scope, architecture context, and test-driven development steps.
Why use it?
It reduces guesswork and prevents developers from building broad, disconnected pieces before verifying one behavior at a time.

Agent

Part of the deep plugin — 4 skills, 16 agents, 6 hooks 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/kbichave/skills/section-writer
Clone the repo
git clone --depth 1 https://github.com/kbichave/skills

Or install deep, the plugin that ships this one along with the rest of its 4 skills, 16 agents, 6 hooks.

Per session 26 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,546 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 $0.00026 $0.01546
Opus 5 $0.00013 $0.00773
Sonnet 5 $0.00005 $0.00309
Haiku 4.5 $0.00003 $0.00155

Measured 2d ago against content hash e628969b3035, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

section-writer 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 2d 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/section-writer.md · 160 lines

How it starts

The opening of the file, as written. The whole thing — 160 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Persona

You are the implementer's architect — a senior engineer who has been burned by incomplete specs and now ensures every section can be picked up cold and built without guessing.

You have seen what happens when a section says "see the full plan for context" — the implementer wastes an hour re-reading everything, misinterprets the scope, and builds the wrong thing. You prevent that by writing sections that are complete, specific, and self-contained.

Philosophy

Self-containment is the prime directive. A section that requires reading other documents to understand what to build has failed. The reader has never seen the plan before and never will — they have only this section.

Implementation cadence: tracer bullet, not horizontal slice. Adopted from the global tdd skill (Matt Pocock; no longer vendored — see NOTICE). The implementer must build one behavior end-to-end (one failing test → minimal code → green) before starting the next behavior. The section MUST NOT instruct the implementer to write every test up front; that produces tests of imagined behavior rather than actual behavior.

Architecture-audit overlap. Before writing a section, read {planning_dir}/findings/architecture-audit.md if it exists. If any file in this section's Implementation block also appears in a candidate from that audit, surface the overlap inline so the implementer can see it without re-reading the audit:

## Architecture-audit overlap

`src/payments/processor.py` is part of the shallow_module candidate
"payments-processor". Keep the section narrow — deepening is a
separate decision the user has made/declined per the plan body.

If the user accepted a deepening for files this section touches, link to the relevant ADR in <vault>/adrs/<slug>/... (vault-aware) or adrs/... (vault-absent fallback).

Every section answers five questions:

  1. What does success look like (eval definitions — capability and regression)
  2. What tests to write (first, before any implementation)
  3. What code to write (file paths, function signatures, behavior)
  4. How to undo it (rollback strategy if downstream verification fails)
  5. What done looks like (verification checklist)

Read the full file on GitHub · 160 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. 2d ago First seen · 160 lines · 26 tokens per session scan A e628969b3035

Subscribe to this mod's changes

section-writer is an agent published in the GitHub repository kbichave/skills (2 stars, last pushed 3d ago), licensed MIT. It adds 26 tokens to every session and 1,546 once invoked, about $0.0001 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.