known-issue-search
Search bug trackers, release notes, and source history before concluding misconfiguration or expected behavior on any bug-shaped ticket.
Skill by c-mongan
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/c-mongan-known-issue-search
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: design, data, productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/known-issue-search
What's inside
- Skill name
known-issue-search- Description
- Search bug trackers, release notes, and source history before concluding misconfiguration or expected behavior on any bug-shaped ticket. Use this skill any time a ticket reports something that worked once and now does not, an upload/build that logs success but the product cannot use, UI state that disagrees with stored data, or any silent no-op / dropped / truncated / zero-result outcome that should not happen.
- License
- MIT
- Instructions
- 105 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Known-Issue Search
Use this skill for every bug-shaped ticket before saying "expected behavior", "misconfiguration", "not a bug", or "no matching issue". One empty search is never evidence that no bug exists.
## Bug-shape signals
A ticket is bug-shaped if it reports any of:
- A feature worked once and now does not.
- An upload, build, or sync logs success but the product cannot use the result.
- UI state disagrees with stored data.
- Aggregations disagree with raw rows.
- Fresh install or documented setup fails out of the box.
- Customer cites silent no-ops, dropped requests, truncated values, or zero-result states that should not happen.
## Search matrix
Run all four query types. Cite each in the report.
### 1. Hybrid discovery
Catches paraphrased reports that lexical search misses.
```bash
gh api '/search/issues?q=repo:OWNER/REPO+<url-encoded symptom phrases>&search_type=hybrid&per_page=10' \
--jq '.items[] | {number,title,state,html_url,created_at,updated_at}'
```
If the product has multiple SDK repos, run hybrid against each likely repo in parallel.
### 2. Broad lexical noun search
2–4 short noun / API / log phrases, not the full customer sentence.
```bash
gh search issues '<short phrase>' --repo OWNER/REPO --match title,body \
--json number,title,state,createdAt,updatedAt,url --limit 20
```
Examples of good phrases: `distinct ID`, `world map`, `displaySurvey`, `stdin`, `Found 0 chunks to upload`, `silent failure`, `symbol set`, `hobby docker compose`.
### 3. Exact surface search
Quoted method names, CLI flags, exact log fragments, visible UI copy.
```bash
gh search issues '"displaySurvey"' --repo OWNER/REPO --match title,body --json number,title,state,url --limit 20
```
### 4. Recent issue sweep
Catches regressions where keywords have not yet stabilized.
```bash
gh search issues '<broad term>' --repo OWNER/REPO --sort updated --limit 20 \
--json number,title,state,createdAt,updatedAt,url
```
### No shell? Use the search API over GET
The shipped `support-triage-agent` has no shell tool. Run the same queries by fetching the GitHub search API with WebFetch, for example `https://api.github.com/search/issues?q=repo:OWNER/REPO+%22exact+phrase%22&sort=updated&per_page=20`. Unauthenticated search is rate-limited; record a gap rather than retrying more than twice.
### Offline demo: `mock-sources/`
If the working directory contains `mock-sources/issues/`, run all four query types there with Grep: hybrid (synonyms and paraphrases), lexical (short nouns), exact (quoted strings) and recent (compare `updated:` front matter with the ticket timeframe). Cite issues by path, for example `mock-sources/issues/BEACON-142.md`.
## Footguns
- Do not run `gh search issues --state all`. `all` is not a valid state. Omit `--state`, or run open and closed separately.
- An empty over-specific query is a dead end, not a conclusion. Broaden terms.
- If the bug-tracker MCP returns full issue bodies, summarize only sanitized facts in the report — bodies often contain customer-identifying details.
- For source-code citations, use commit-pinned permalinks, not `blob/main`. The blob SHA from the contents API is not a commit SHA — fetch it separately:
```bash
gh api repos/OWNER/REPO/commits/$(gh api repos/OWNER/REPO --jq '.default_branch') --jq '.sha'
```
## Cross-repo correlation
When a bug surfaces in one SDK or component, automatically check related repos. Issues often span repositories — a web-SDK bug may have a backend cause; a mobile-SDK bug may also affect another platform. After finding a match in one repo, fire a parallel search in correlated repos and note any cross-repo matches in the evidence.
## Output
```markdown
## Known-Issue Search
| # | Query type | Repo | Query | Best Candidate | Conclusion |
|---|---|---|---|---|---|
| 1 | Hybrid | OWNER/REPO | … | … | match / rejected / no match |
| 2 | Lexical | OWNER/REPO | … | … | match / rejected / no match |
| 3 | Exact | OWNER/REPO | … | … | match / rejected / no match |
| 4 | Recent | OWNER/REPO | … | … | match / rejected / no match |
**Matching issue / release:** …
**Closest rejected candidates:** … (with one-line reasons)
**Search gaps:** … (e.g., MCP unavailable, rate-limited, repo missing)
```
## Confidence judgment
- Exact symptom + matching version range usually supports **Likely** confidence.
- Exact symptom + direct project/customer evidence can support **Confirmed** confidence.
- Similar symptom without version or config verification stays **Suspected** or **Likely** depending on evidence strength.
- Asserting "no match found" without all four query types caps confidence at **Low (40%)**.From c-mongan/support-triage-agent-template, MIT.
More from c-mongan/support-triage-agent-template
- 0.565score
log-evidence
c-mongan
Read customer-supplied logs, stack traces, HAR files and browser console output as evidence.
SkillAFreshness: AClaude Code +9 - 0.565score
redaction
c-mongan
Strip secrets, tokens, raw person properties, private URLs, and internal admin links before saving any triage output to disk or copying into customer-facing…
SkillAFreshness: AClaude Code +9 - 0.564score
escalation
c-mongan
Package support investigations into engineering-ready escalation briefs.
SkillAFreshness: AClaude Code +9 - 0.564score
response-drafting
c-mongan
Draft customer-facing support responses from verified triage findings.
SkillAFreshness: AClaude Code +9 - 0.564score
triage-report
c-mongan
Produce a structured, evidence-graded support triage report.
SkillAFreshness: AClaude Code +9 - 0.561score
kb-article
c-mongan
Turn a resolved or well-understood triage report into a draft knowledge-base article (how-to, troubleshooting, known issue or FAQ).
SkillAFreshness: AClaude Code +9 - 0.561score
reproduction-steps
c-mongan
Write a minimal, numbered reproduction for a support ticket, with environment, expected and actual results, so an engineer can confirm a suspected bug.
SkillAFreshness: AClaude Code +9 - 0.560score
support-triage
c-mongan
Run the read-only support triage workflow on a ticket and save an evidence-graded triage report plus a draft customer response under reports/.
SkillAFreshness: AClaude Code +9 - 0.558score
ticket-intake
c-mongan
Normalize a raw support ticket into structured investigation inputs.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 7, 2026 (21 days ago)
- Last release
- Oct 7, 2026 (21 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from c-mongan
- 0.565score
log-evidence
c-mongan
Read customer-supplied logs, stack traces, HAR files and browser console output as evidence.
SkillAFreshness: AClaude Code +9 - 0.565score
redaction
c-mongan
Strip secrets, tokens, raw person properties, private URLs, and internal admin links before saving any triage output to disk or copying into customer-facing…
SkillAFreshness: AClaude Code +9 - 0.564score
escalation
c-mongan
Package support investigations into engineering-ready escalation briefs.
SkillAFreshness: AClaude Code +9 - 0.564score
response-drafting
c-mongan
Draft customer-facing support responses from verified triage findings.
SkillAFreshness: AClaude Code +9 - 0.564score
triage-report
c-mongan
Produce a structured, evidence-graded support triage report.
SkillAFreshness: AClaude Code +9 - 0.561score
kb-article
c-mongan
Turn a resolved or well-understood triage report into a draft knowledge-base article (how-to, troubleshooting, known issue or FAQ).
SkillAFreshness: AClaude Code +9
Cite this page
known-issue-search skill. Agent Mart. https://agentmart.dev/skills/c-mongan-known-issue-search (last updated 2026-10-24)