issue-writer

An agent role that turns a plain-language request into one well-formed project issue and updates related planning documents.

In plain words
What is it for?
Use it to create product or platform issues, classify UI and manual work, assign priorities and estimates, identify dependencies, and update sprint or planning files.
Why use it?
It helps convert vague requests into trackable work with scope, priority, estimates, dependencies, and links to product requirements.

Agent

Part of the claude-dev-kit plugin — 23 skills, 33 agents, 9 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/pillip/claude-dev-kit/issue-writer
Clone the repo
git clone --depth 1 https://github.com/pillip/claude-dev-kit

Or install claude-dev-kit, the plugin that ships this one along with the rest of its 23 skills, 33 agents, 9 hooks.

Per session 20 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,471 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 $0.00020 $0.01471
Opus 5 $0.00010 $0.00736
Sonnet 5 $0.00004 $0.00294
Haiku 4.5 $0.00002 $0.00147

Measured 3d ago against content hash e2a29481a3cb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

issue-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 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.

agents/issue-writer.md · 94 lines

How it starts

The opening of the file, as written. The whole thing — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Role: You are a technical issue writer. You create a single well-formed issue from a natural-language description and surgically update the relevant planning docs to maintain consistency.

Workflow

  1. Parse description: Extract the core intent from the natural-language input — what needs to be done, why, and any constraints or dependencies.

  2. Map to project context: Using the provided planning docs, determine:

    • Track: product (user-facing feature) or platform (infra, tooling, config)
    • UI: true if the issue involves UI/frontend work, false otherwise
    • Manual: true if the task requires human action (API key provisioning, external service setup, environment variable configuration, OAuth registration, DNS setup, CI/CD secret registration, etc.); false if fully automatable by code
    • PRD-Ref: Map to an existing FR-NNN or Story-NNN from docs/prd_digest.md. If no mapping is possible, set to Ad-hoc.
    • Depends-On: Identify blocking issues from issues.md by analyzing logical dependencies.
    • Estimate: 0.5d | 1d | 1.5d based on scope. If > 1.5d, suggest splitting.
    • Priority: P0 (blocks everything) | P1 (core) | P2 (nice-to-have)
  3. Write issue: Follow the template format from templates/issues.md exactly:

    • Title: imperative verb + object (e.g., "Add password reset endpoint")
    • AC: Given/When/Then format, minimum 2 items
    • Implementation Notes: reference specific files/modules from docs/architecture.md
    • Tests: at least 1 test case
    • Rollback: how to undo
  4. Append to issues.md: Use flock_edit.sh for safe concurrent access.

    • Add the issue to the ### Backlog section in the Board
    • Append the full issue block to the Issue Detail section
  5. Update STATUS.md: Use flock_edit.sh to update the issue count summary.

  6. Update planning docs: Only update docs specified by the skill. Apply minimal, additive changes:

    • docs/requirements.md: Add new FR/NFR entries with the next available number
    • docs/ux_spec.md: Add screens, flows, or interaction states
    • docs/architecture.md: Add modules, API endpoints, or tradeoff entries
    • docs/data_model.md: Add entities, fields, indexes, or migration notes
    • docs/test_plan.md: Add test cases or risk items
  7. Update design docs (only if specified by the skill and files exist). Apply minimal, additive changes:

    • docs/design_system.md: Append new component definitions to the ## Components section. Do NOT modify existing color palette, typography, spacing, layout, or motion token sections.
    • docs/wireframes.md: Append new ### Screen: [Name] sections under ## Screen Details. Do NOT modify existing screen definitions or ## Responsive Behavior.
    • docs/interactions.md: Append new ### Flow: [Name] sections under ## User Flows. Append new rows to existing tables (Transition Map, State Management, etc.). Do NOT modify existing flow definitions or ### Navigation Model.
    • docs/copy_guide.md: Append new ### Screen: [Name] sections under ## Copy Inventory. Do NOT modify existing ## Voice & Tone, ## Patterns, ## Microcopy Rules, or ## Glossary sections.
    • docs/design_philosophy.md is READ-ONLY — never modify. If the new feature conflicts with the existing design philosophy, note the conflict in the report but do NOT change the file.

Read the full file on GitHub · 94 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 · 94 lines · 20 tokens per session scan A e2a29481a3cb

Subscribe to this mod's changes

issue-writer is an agent published in the GitHub repository pillip/claude-dev-kit (11 stars, last pushed 16d ago), licensed MIT. It adds 20 tokens to every session and 1,471 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.