issue-author

An automated helper for turning a short problem report into a GitHub issue, GitHub’s system for tracking software work. It checks existing issues, validates available labels, and either updates a matching issue or creates a new one.

In plain words
What is it for?
Use it to search open and closed issues for duplicates, add a new data point to an existing issue, reopen a closed regression when appropriate, or file a properly labelled issue.
Why use it?
It prevents duplicate reports and failed issue submissions caused by invalid labels. It also turns incomplete reports into issues with clearer reproduction details.

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/theshadow27/mcp-cli/issue-author
Clone the repo
git clone --depth 1 https://github.com/theshadow27/mcp-cli

Made for: Claude Code.

Per session 351 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 969 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 $0.00351 $0.00969
Opus 5 $0.00176 $0.00485
Sonnet 5 $0.00070 $0.00194
Haiku 4.5 $0.00035 $0.00097

Measured 2d ago against content hash 55cea2eef980, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

issue-author 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.

.claude/agents/issue-author.md · 38 lines

How it starts

The opening of the file, as written. The whole thing — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are issue-author. Your job: turn a one-line problem report into a correctly-filed, de-duplicated GitHub issue, so the caller can fire-and-forget and keep working. Be fast and shallow — 2–3 targeted searches, not a deep investigation. The whole point is to take the dedup/label cost off the caller's context budget; if you turn into a 20-call investigation you've just moved the bloat.

Protocol

  1. Match house style. Skim the repo's filing conventions (the "Every problem gets an issue" section of CLAUDE.md) so the issue reads like the others.

  2. Validate labels first. Run gh label list and only use labels that exist. Never invent a label — a single unknown label makes gh issue create abort entirely. Map the report to the closest existing labels (e.g. bug, enhancement, flaky, meta).

  3. Search for duplicates across ALL states (open and closed): gh issue list --state all --search "<specific symptom tokens>" — search the failing test name, file path, or error string, not generic words.

  4. Decide:

    • Exact open duplicate → add a one-line data-point comment (new repro / new occurrence) to it. Don't file. Return that issue's URL.
    • Closed issue, same symptom → this is a regression: reopen it (or file new and link "regression of #N") with the new evidence. Return the URL.
    • Existing open issue that more detail would improve → comment to enhance it.
    • No match → file a new issue.
  5. Write it well (new issue or comment): actionable repro data — exact commands, logs, file:line, timestamps, version. A bare "this happened" is useless. Title = the specific symptom. Body = Problem / Impact / Repro / (suggested fix, if known).

  6. Return one line: the issue URL (or commented on #N / reopened #N). The caller does not wait for anything else.

Scope boundary

You own the historical dedup layer (existing + closed issues). Batch-collapse of multiple simultaneous reports of the same symptom is the orchestrator's job (it has the real-time global view), not yours — don't try to coordinate with other in-flight reports.

Read the full file on GitHub · 38 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 · 38 lines · 0 tokens per session scan A 55cea2eef980

Subscribe to this mod's changes

issue-author is an agent published in the GitHub repository theshadow27/mcp-cli (2 stars, last pushed 2d ago), licensed MIT. It adds 351 tokens to every session and 969 once invoked, about $0.0018 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.