issue-reporting — Claude Code skill: what it does, how to install
issue-reporting turns a complaint about a GTM skill into a GitHub issue maintainers can act on: it collects the details, sorts the failure type and prefills the bug template.
MIT 118 installs ★ 114 updated 2026-09-18
npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a claude-codeWhat issue-reporting does
The skill takes a complaint about another skill in the claude-gtm-plugin pack and writes it up as a GitHub issue. It asks which skill was involved, the prompt you sent, the output you got and the output you wanted. A vague complaint gets a follow-up question rather than a filed report.
It then checks the project: whether strategy/brand.md, about/me.md and CLAUDE.md are there, which version sits in .claude-plugin/plugin.json, and how the plugin was installed. The case lands in one bucket — wrong advice, unwanted activation, missing context, a crash. You get a prefilled bug_report.yml link plus a markdown body to paste by hand.
When the brand files are absent, it points you at /bootstrap first. Generic advice usually means missing context, not a broken skill. Patching is out of scope: this one writes the report, a maintainer fixes the code.
What it helps with
- file a GitHub issue about a skill that gave wrong advice
- tell a missing brand-context problem from an actual skill bug
- capture the prompt and the bad output before they get lost
- pin down why a skill never triggered on your phrase
- attach plugin version and install method to the report
Example prompts for the agent
- The copywriting skill gave me wrong positioning advice, help me file an issue
- seo-and-aeo-strategy never activated on my prompt, let's write up a bug report
- Something is broken: the skill ignored my brand context, start an issue on GitHub
How to install issue-reporting
1Skills CLI (Vercel)
npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a claude-code2Manually: put the files in place
~/.claude/skills/issue-reporting/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/issue-reporting/
1Skills CLI (Vercel)
npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a codex2Manually: put the files in place
.agents/skills/issue-reporting/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a cursor2Manually: put the files in place
.cursor/skills/issue-reporting/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install manojbajaj95/claude-gtm-plugin skills/issue-reporting2Manually: put the files in place
.github/skills/issue-reporting/SKILL.md1Skills CLI (Vercel)
npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a gemini-cli2Platform CLI
gemini skills install https://github.com/manojbajaj95/claude-gtm-plugin --path skills/issue-reportingManually: put the files in place
.devin/skills/issue-reporting/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about issue-reporting
How do I install issue-reporting in Claude Code?
Run npx skills add manojbajaj95/claude-gtm-plugin --skill issue-reporting -a claude-code. The skill file ends up at ~/.claude/skills/issue-reporting/SKILL.md.
What does issue-reporting do?
It walks you from a complaint to a finished issue: skill name, prompt, actual versus expected output, a check of the context files and plugin version, then the report text.
Why does it suggest /bootstrap first?
Without strategy/brand.md and about/me.md the other skill has nothing to tailor advice to. If the recommendation is still wrong after the context is set up, that is a real bug worth filing.
Source text
SKILL.md · MIT
---
name: issue-reporting
description: "Helps users report issues with GTM skills — incorrect advice, broken activation, missing context, or unexpected behavior. Guides through gathering the right details (skill name, prompt used, actual vs expected output, context files present) and generates a properly formatted GitHub issue. Use when users say: 'this skill gave wrong advice', 'report a bug', 'file an issue', 'skill isn't working', 'wrong output', 'skill didn't activate', 'bad recommendation', 'incorrect behavior', or 'something is broken'."
---
# Issue Reporting
## Workspace Context
Read bootstrap context before asking questions: `strategy/brand.md` for brand, audience, offer, channels, tools, constraints, and metrics; `about/me.md` for personal voice; `content/ideas.md` and `content/calendar.md` for content planning. Use legacy product-marketing context files only as fallback. Save generated drafts to `content/<platform>/drafts/YYYY-MM-DD_short-topic-slug.md`, and route durable learnings back to `strategy/brand.md`, `about/me.md`, or `content/ideas.md`.
## Operating Contract
This skill is self-contained for its frontmatter scope: use its local instructions, references, scripts, and assets as the playbook; ask only for missing task-specific inputs; hand off to adjacent skills instead of expanding scope; and return an actionable artifact, decision, plan, draft, or diagnostic.
Help users report problems with GTM skills clearly and completely so maintainers can fix them fast.
## When This Skill Activates
- User says a skill gave wrong or harmful advice
- User says a skill isn't activating when it should
- User says their brand context wasn't applied
- User says output format is wrong or unexpected
- User wants to report a bug or file an issue
- User says "something is broken" about a skill
## Step 1 — Identify the Problem
Ask the user what went wrong. Determine:
1. **Which skill** — get the exact skill name (e.g. `seo-and-aeo-strategy`, `copywriting-core`)
2. **What they asked** — the exact prompt or task they gave Claude
3. **What happened** — the actual output or behavior
4. **What they expected** — what should have happened instead
If the user is vague ("it gave bad advice"), ask specifically what was incorrect and why.
## Step 2 — Gather Context
Check the user's project for context that helps diagnose the issue:
### Context files
Check if these exist and note their presence:
- `strategy/brand.md` — brand positioning, messaging, audience
- `about/me.md` — personal voice, writing style
- `CLAUDE.md` — AI instructions
Missing context files are the **most common cause** of "wrong advice" bugs. If `strategy/brand.md` or `about/me.md` are missing, tell the user:
> "The skill may have given generic advice because your brand context files are missing. Run `/bootstrap` to set them up. If the skill still gives wrong advice after that, then it's a real bug worth reporting."
### Plugin version
Check `.claude-plugin/plugin.json` for the current version number.
### Installation method
Ask how they installed the plugin (Claude Code, Cowork, Skills CLI, git clone, submodule).
## Step 3 — Classify the Issue
Categorize into one of these types:
| Type | Description | Common cause |
|---|---|---|
| Wrong advice | Skill produced incorrect or harmful recommendations | Outdated framework, wrong domain assumptions |
| Not activating | Skill should have triggered but didn't | Description mismatch, trigger phrase gap |
| Wrong activation | Skill triggered when it shouldn't have | Overly broad description |
| Missing context | Brand/voice not applied to output | Context files missing or not read |
| Bad format | Output structure is wrong | Skill template issue |
| Crash/error | Skill fails mid-execution | Syntax error in skill, bad reference path |
## Step 4 — Generate the Issue
Compose a GitHub issue using the bug report template. Format it so the user can copy-paste or click through:
**Direct link** (pre-fills the template):
```
https://github.com/manojbajaj95/claude-gtm-plugin/issues/new?template=bug_report.yml
```
**Formatted issue body** (if they prefer to paste manually):
```markdown
**Skill**: [skill-name]
**Type**: [category from Step 3]
**Plugin Version**: [version]
**Installation**: [method]
### What I asked
[exact prompt]
### What happened
[actual output — include relevant excerpt]
### What I expected
[expected behavior]
### Context files present
- [ ] strategy/brand.md
- [ ] about/me.md
- [ ] CLAUDE.md
### Steps to reproduce
1. [step-by-step if applicable]
### Impact
[Critical / High / Medium / Low]
```
## Step 5 — Offer Next Steps
After generating the issue:
1. Give them the direct link to file it: `https://github.com/manojbajaj95/claude-gtm-plugin/issues/new?template=bug_report.yml`
2. If the problem is missing context files, suggest running `/bootstrap` first
3. If it's clearly a skill bug, encourage them to file it — it helps improve the plugin for everyone
4. Thank them for reporting — contributors who file clear issues are invaluable
## Important
- Never dismiss user frustration — if they think the advice is wrong, take it seriously
- Don't try to fix the skill yourself — this skill is for *reporting*, not patching
- Be specific in the issue — vague reports like "it didn't work" are hard to act on
- Include the actual output when possible — maintainers need to see what went wrong
© manojbajaj95 · MIT · manojbajaj95/claude-gtm-plugin