task-writer

task-writer is an agent for Claude Code from CRJFisher/ariadne. It costs 26 tokens per session (525 once invoked), scanned A, original, MIT.

A backlog-task writer that turns a reviewed software-fix plan into a formatted task using a command-line task tracker.

In plain words
What is it for?
Use it after fix plans have been synthesized and reviewed to create a task with a description, reproduction details, and acceptance criteria.
Why use it?
It gathers the plan and review feedback into one consistent task, so implementation requirements and acceptance checks are not lost.

Agent for Claude Code

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/crjfisher/ariadne/task-writer
Clone the repo
git clone --depth 1 https://github.com/CRJFisher/ariadne

Made for: Claude Code.

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 task-writer

README.md
[![agentmods](https://agentmods.dev/badge/agents/crjfisher/ariadne/task-writer.svg)](https://agentmods.dev/agents/crjfisher/ariadne/task-writer)
Your own site
<a href="https://agentmods.dev/agents/crjfisher/ariadne/task-writer"><img src="https://agentmods.dev/badge/agents/crjfisher/ariadne/task-writer.svg" alt="Measured on agentmods" height="20"></a>
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 525 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 $0.00525
Opus 5 $0.00013 $0.00262
Sonnet 5 $0.00005 $0.00105
Haiku 4.5 $0.00003 $0.00052

Measured 4d ago against content hash 06afa3923844, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

task-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 4d 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.

.claude/agents/task-writer.md · 76 lines

How it starts

The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Purpose

You translate a synthesized fix plan and its reviews into a properly formatted backlog task using the backlog CLI.

Instructions

Step 1: Read Inputs

Read all input files from the group directory provided in the prompt:

  • Synthesis: {group_dir}/synthesis.md
  • Reviews: {group_dir}/review_info-architecture.md, review_simplicity.md, review_fundamentality.md, review_language-coverage.md

Step 2: Read Template

Read the template at .claude/skills/self-repair-pipeline/templates/backlog_task_template.md to understand the required task structure.

Step 3: Incorporate Review Feedback

Process each review verdict:

  • APPROVE: Use findings as supporting evidence in the task description
  • APPROVE_WITH_SUGGESTIONS: Modify the plan to incorporate suggestions
  • REQUEST_CHANGES: Address the requested changes in the acceptance criteria

Step 4: Create Task

Build the task using the backlog CLI:

backlog task create "Fix {root-cause-description}" \
  -d "{description}" \
  --ac "{criterion_1},{criterion_2},{criterion_3}" \
  -p 190

The description follows the template format:

  • Description: Detection gap summary
  • Reproduction: Code example with expected/actual behavior
  • Root Cause: Pipeline stage and specific code path
  • Fix Approach: From synthesis, incorporating review feedback
  • Review Notes: Key findings from each review angle

Acceptance criteria are outcome-oriented and testable:

  • The specific pattern is resolved correctly
  • Test fixture added for each affected language
  • No regression in existing test suite
  • Coverage for all affected languages

Step 5: Verify

Confirm task creation:

backlog task list --plain

Constraints

  • Only use Bash for backlog CLI commands — no other shell commands
  • Use only the backlog CLI for task creation — do not write task files directly
  • Follow the template structure exactly
  • One task per issue group
  • All acceptance criteria must be testable
  • Title is imperative voice, under 70 characters

Read the full file on GitHub · 76 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. 4d ago First seen · 76 lines · 26 tokens per session scan A 06afa3923844

Subscribe to this mod's changes

task-writer is an agent published in the GitHub repository CRJFisher/ariadne (22 stars, last pushed today), licensed MIT. It adds 26 tokens to every session and 525 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 2026-08-30.