issue
Use GitHub issues as persistent cross-session resolution memory. Create, update, and re-read issues that are self-contained and re-readable cold, recording the hypothesis and the resolution. Triggers on "track this in an issue", "create an issue I can resume", "resume issue", "is
Install
npx skills add https://github.com/mirkobozzetto/arsenal/tree/main/plugins/issue/skills/issue
claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install mirkobozzetto-arsenal@llmmart
git clone https://github.com/mirkobozzetto/arsenal.git
The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole mirkobozzetto/arsenal collection as a plugin from our marketplace. Git is the plain clone.
Skill manifest
Issue memory
Use the authenticated gh CLI. Read the specific issue and comments before updating it; do not query unrelated issues. Capture the problem, evidence, attempts, remaining action and verification so the issue can be resumed cold. Use references/issue-template.md when creating one.
Confirm before creating or closing an issue. Append requested progress as comments instead of overwriting the body. Preserve the existing claude-memory label and Pickup Directive contract. No code changes, commits or agents.
Resume reconstructs the current state from the issue, not stale memory. A solved issue does not trigger another diagnosis. List returns concise open items and stops. If the controlled shell has no external networking, report that authorization/capability boundary instead of claiming the GitHub action happened or attempting an execution escape.
Arsenal handoff
Return issue context and remaining actions to Arsenal when it called this skill. A request to fix the issue can continue through Arsenal's implementation route; reading or resuming context alone does not authorize code changes. Issue creation and closure retain their confirmation gates.
Execution policy
Work solo. Ask before any subagent or reviewer, even in auto mode. Explain the independent scope and expected benefit first. No hidden advisor, nested delegation, model retuning, repeated successful checks, or progress spam. Use existing context before asking questions. Stop when the requested result is delivered. User stops and scope changes override pending steps.
Files (arsenal)
-
references
-
issue-template.md 1.1 KB
# Issue body template Render this as the GitHub issue body on `create`. Every section is required; the **Pickup Directive** is what lets a future session resume cold. Prose in English; identifiers in English. ```markdown ## Problem <what is broken / what must be solved, observable symptom, where it shows up> ## Pickup Directive > Self-contained. A future session reads ONLY this section + comments to resume, with no prior chat. - Context: <repo, branch, files/areas involved> - Current state: <what is done / not done> - Next step: <the concrete action to resume> - How to verify: <command / test / success signal> ## Hypothesis <suspected cause, and why> ## Attempts - [ ] <attempt + result> (updated via `gh issue comment`) ## Resolution <filled in at close: what fixed it, and why it works> ``` ## Conventions - Title: concise, problem-first (e.g. `Auth: token expiry off-by-one on refresh`). - Label: `claude-memory` (filter for `resume` / `list`). - Progress goes in **comments**, not by editing the original body: the body stays the stable Pickup Directive. - On close: append a `Resolution:` comment, then `gh issue close <N> --reason completed`.
-
-
SKILL.md 1.7 KB
--- name: issue description: Create, update or resume a GitHub issue as durable resolution memory when requested. argument-hint: "[create|update|resume|list] [#N | problem]" --- # Issue memory Use the authenticated gh CLI. Read the specific issue and comments before updating it; do not query unrelated issues. Capture the problem, evidence, attempts, remaining action and verification so the issue can be resumed cold. Use references/issue-template.md when creating one. Confirm before creating or closing an issue. Append requested progress as comments instead of overwriting the body. Preserve the existing claude-memory label and Pickup Directive contract. No code changes, commits or agents. Resume reconstructs the current state from the issue, not stale memory. A solved issue does not trigger another diagnosis. List returns concise open items and stops. If the controlled shell has no external networking, report that authorization/capability boundary instead of claiming the GitHub action happened or attempting an execution escape. ## Arsenal handoff Return issue context and remaining actions to Arsenal when it called this skill. A request to fix the issue can continue through Arsenal's implementation route; reading or resuming context alone does not authorize code changes. Issue creation and closure retain their confirmation gates. ## Execution policy Work solo. Ask before any subagent or reviewer, even in auto mode. Explain the independent scope and expected benefit first. No hidden advisor, nested delegation, model retuning, repeated successful checks, or progress spam. Use existing context before asking questions. Stop when the requested result is delivered. User stops and scope changes override pending steps.
Comments (0)
Sign in to join the conversation.
Reviews (0)
No reviews yet.
No comments yet.