apollo-issue-review_skill

This skill analyzes Apollo ecosystem issues using a classify-first workflow to draft maintainer-grade replies with actionable next steps.
  • Python

2.5k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill apollo-issue-review

  • _meta.json293 B
  • SKILL.md10.1 KB

Overview

This skill reviews issues across the Apollo ecosystem with a classify-first workflow and drafts maintainer-grade replies. It classifies issues as behavior/regression or consultative/support, performs the appropriate validation (reproduce or evidence-check), and produces a concise, actionable reply plus triage guidance. The output is ready for maintainer review and can be held as a draft or posted after confirmation.

How this skill works

The skill ingests issue metadata (repo, issue number, title/body/comments, labels, desired outcome) and decides issue type up front. For behavior/regression it builds minimal reproductions or unit-test checks and records exact observed outputs; for consultative/support it performs a targeted repository evidence scan (controllers, docs, OpenAPI scripts, pom.xml). It then drafts a reply that states the validation result, current support boundaries, exact file/API paths or commands, and next steps or contribution guidance.

When to use it

  • Triage reported runtime bugs or regressions where a concrete reproduction is expected
  • Answer API/docs/capability questions that need proof from code or docs
  • Draft maintainer replies that must be concise, non-speculative, and action-focused
  • Prepare an implementation handoff for contributors after a confirmed gap or agreed design
  • Avoid duplicate long replies when prior comments already solved the ask

Best practices

  • Classify first: reproduction for behavior, evidence-check for consultative asks
  • Use canonical Apollo module names and correct misnomers succinctly
  • Record exact files/paths/commands searched and observed outputs — avoid speculative root causes
  • If reproducible and conclusion is stable, do not ask for extra logs
  • When a community member volunteers to implement, acknowledge then give feasibility boundaries and concrete refinements
  • Always include a publish gate: draft only by default, post only after explicit approval

Example use cases

  • User reports an agent crash: reproduce locally, confirm stack trace, propose workaround and label bug
  • User asks whether an OpenAPI endpoint exists: scan controllers, docs, pom.xml and provide exact path or alternative
  • Contributor offers a patch: evaluate feasibility, specify files/tests to touch and acceptance criteria
  • User reports degraded behavior across versions: reproduce on specified version and clarify whether regression analysis is requested
  • Clarify terminology in an issue that used non-canonical module names and provide corrected module references

FAQ

Provide repo, issue_number, and issue_context (title/body/comments). If missing, the skill asks one brief clarification before continuing.

Will the skill post replies automatically?

No. By default it generates a draft and asks for explicit confirmation before posting to the issue.

How does it handle duplicate prior answers in the thread?

If an existing comment already answers the ask, the skill posts only a concise supplement or correction instead of repeating the full reply.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational