AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified Apache-2.0 Self-run

Issues

skill-loomantix-claude-platform-issues · by loomantix

GitHub issue workflow — ready queue with dependency resolution, claim, start, close, link

No reviews yet
0 installs
22 views
0.0% view→install

Install

$ agentstack add skill-loomantix-claude-platform-issues

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-loomantix-claude-platform-issues)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
18d ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Issues? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

/issues

Thin workflow over gh issue with a smart ready query that parses Blocked by #N / Depends on #N from issue bodies to compute dependency-free work. (Blocks #N is written as the reciprocal side by /issues link, but isn't parsed as a blocker itself — the authoritative direction is Blocked by.)

Arguments: $ARGUMENTS

Dispatch on the first word of $ARGUMENTS. If no verb is given, default to ready.


Permission hygiene

Never use heredocs in gh commands — the auto-approval regex can't match multiline commands, which causes permission prompts. Always write multiline content to a temp file first:

cat > /tmp/issue-body.md  --body-file /tmp/issue-body.md

Apply to: gh issue create, gh issue edit, gh issue comment.


ready

Show open issues with no open blockers, sorted by priority.

./.claude/skills/issues/scripts/ready.py

Flags (all optional):

  • --mine — only issues assigned to me
  • --unassigned — only unassigned issues
  • --agent — only issues labeled dev: agent
  • --priority critical|high|medium|low
  • --area (matches area: label — e.g., backend, frontend, mobile, packages)
  • --limit N (default 20)
  • --json — machine-readable

Exclusion rules:

  • Label status: blocked
  • Label status: on-staging — fix merged to a staging/integration branch, awaiting release/promotion (done, pending; an opt-in convention — no-op in repos that don't apply it)
  • Any agent-bail:* label — explicitly excluded by backlog refinement or a prior loop run, even if a stale dev: agent label remains
  • Body contains Blocked by #N or Depends on #N where #N is still open
  • Targeted by a closing reference from an open PR, or from a PR merged in the last 30 days (via closingIssuesReferences, with a closing-keyword body fallback) — keeps issues already fixed as a PR side-item or by an in-review PR out of the queue, including done-on-integration issues that a non-default-branch merge never auto-closed

The --agent / --priority / --area flags work via standard label conventions (dev: agent, priority: , area: ). Repos that don't use those labels will simply get an empty result for those filters — the script doesn't enforce a label scheme, it just queries one when asked.


show \

gh issue view 

Surface dependency refs explicitly (useful for triage):

gh issue view  --json body --jq '.body' | grep -iE '^[[:space:]]*[-*]?[[:space:]]*(blocked by|blocks|depends on)[:\s]+#[0-9]+' || echo "(no dependency refs)"

claim \

gh issue edit  --add-assignee @me
printf 'Claiming this.\n' > /tmp/issue-comment.md
gh issue comment  --body-file /tmp/issue-comment.md

start \

Claim + create working branch off the repo's default branch. The default branch is auto-detected (main, staging, etc.) via the upstream HEAD ref so this works across consumer repos without per-repo config:

gh issue edit  --add-assignee @me
slug=$(gh issue view  --json title --jq '.title' | tr '[:upper:]' '[:lower:]' | sed 's/[^a-z0-9]/-/g; s/--*/-/g; s/^-//' | cut -c1-50 | sed 's/-$//')
default_branch=$(git symbolic-ref refs/remotes/origin/HEAD 2>/dev/null | sed 's@^refs/remotes/origin/@@')
default_branch="${default_branch:-main}"
git fetch origin "$default_branch"
git checkout -b "fix/issue--$slug" "origin/$default_branch"

Replace ` and the slug interpolation with the real issue number. If git symbolic-ref is unset (rare; happens when the remote was added without --mirror or git remote set-head was never run), the fallback is main. Run git remote set-head origin --auto` once on the affected clone to fix it permanently.


close \ \[msg\]

With a comment:

printf '\n' > /tmp/issue-close.md
gh issue close  --comment "$(cat /tmp/issue-close.md)"

Without:

gh issue close 

Confirm with the user before closing if the issue isn't assigned to them or the close reason isn't obvious from the conversation.


link \ blocks|blocked-by \

Adds dependency refs to both issues so ready sees them regardless of which side you query.

./.claude/skills/issues/scripts/link.py  blocks 
./.claude/skills/issues/scripts/link.py  blocked-by 
  • link A blocks B → writes Blocks #B to A, Blocked by #A to B
  • link A blocked-by B → writes Blocked by #B to A, Blocks #A to B

Refs land under a ## Dependencies section in each body. If a matching ref already exists, the script no-ops that side.


search \

Forward to gh issue list --search:

gh issue list --search "" --limit 20

Useful query fragments: label:"dev: agent", is:open no:assignee, in:title pipeline, author:@me.


Dependency parsing rules

ready recognizes these patterns (case-insensitive, on their own line, optionally bulleted):

  • Blocked by #N
  • Depends on #N

To mark a dependency, prefer /issues link over manual edits — it keeps both issues consistent and under a parseable ## Dependencies section. The label status: blocked is also honored (excludes from ready regardless of body content).


Hard rules

  • Never use heredocs in gh commands — always temp files.
  • Never close someone else's issue without explicit user confirmation.
  • Keep dependency refs under the ## Dependencies section so they stay parseable and don't conflict with prose.
  • If ready returns nothing, don't invent work — report the empty queue and stop.

Source of truth

This skill lives upstream at .claude/skills/issues/. Synced to consumer repos via the sync mechanism. Edits in a consumer will be overwritten on next sync — make changes upstream.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.