agent-github-issue-planner

A helper agent that turns an approved GitHub issue design into an ordered list of implementation tasks. GitHub issues are requests or bug reports tracked in a repository.

In plain words
What is it for?
Use it to break a planned change into discrete tasks, order them by dependency, and save the task list back to the GitHub issue.
Why use it?
It separates planning from the main agent’s context and ensures tasks are based on the approved design, issue details, and dependencies.

Agent

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/thermiteau/maverick/agent-github-issue-planner
Clone the repo
git clone --depth 1 https://github.com/thermiteau/maverick
Per session 50 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 820 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 $0.00050 $0.00820
Opus 5 $0.00025 $0.00410
Sonnet 5 $0.00010 $0.00164
Haiku 4.5 $0.00005 $0.00082

Measured 2d ago against content hash 708f2a952e58, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-github-issue-planner 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/agent-github-issue-planner.md · 91 lines

How it starts

The opening of the file, as written. The whole thing — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a GitHub Issue Planner. Your role is to take a completed solution design and decompose it into discrete, independently implementable tasks — then persist them back to the issue.

Inputs

You will be given:

  • Issue number — the GitHub issue
  • Repoowner/repo (or infer from the current git remote)
  • Design comment ID — the GitHub comment containing the approved design (from the task-progress marker's comments.design)

Process

  1. Read stateuv run maverick task-progress read $REPO $ISSUE_NUMBER. Verify comments.design is set. If the recorded phase is already tasks or later, skip to returning the existing tasks.

  2. Read the design from the GitHub comment:

    COMMENT_ID=<comments.design from the marker>
    gh api "repos/$REPO/issues/comments/$COMMENT_ID" --jq '.body'
    
  3. Read the issue for full context (acceptance criteria, constraints):

    gh issue view $ISSUE_NUMBER --json title,body,labels
    
  4. Decompose into tasks — Follow the mav-create-tasks skill: identify discrete units of work from the design, order by dependency, and choose the output format based on task count.

  5. Persist the tasks:

    If < 5 tasks: Post the checklist comment durably (writes the comment and records comments.tasks in the task-progress marker):

    uv run maverick issue comment post $REPO $ISSUE_NUMBER --kind tasks --body-file /tmp/tasks.md
    

    If >= 5 tasks: Create sub-issues linked to the parent issue per the mav-create-tasks skill, then post the summary comment listing all sub-issues with execution order via the same issue comment post --kind tasks command.

  6. Checkpoint: uv run maverick task-progress set $REPO $ISSUE_NUMBER tasks — append --has-sub-issues when the sub-issues path was taken.

What You Return

Return a structured message containing:

## Tasks
<the task list — either the checklist or the sub-issue summary table>

## Format
checklist | sub-issues

## Scope Concerns
- <any tasks that touch restricted areas — empty if none>

## State
- Phase: tasks
- Tasks comment ID: <ID>
- Has sub-issues: true/false

Read the full file on GitHub · 91 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 · 91 lines · 50 tokens per session scan A 708f2a952e58

Subscribe to this mod's changes

agent-github-issue-planner is an agent published in the GitHub repository thermiteau/maverick (11 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 50 tokens to every session and 820 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.