zentao-bug-triage
Skill3 starsChanged 2 months agoMITView on GitHub
Fetch a lightweight current-project Zentao Bug list, deep-fetch selected Bug IDs with their attachments, or explicitly reconcile exact-target history. Use for 当前bug, 当前分支bug, 抓bug, 看禅道bug, a bare Bug ID that needs fetching, 下载附件, or 回归/状态对账. This skill is read-only in Zentao; use zentao-bug-resolver for remote resolution.
In this repository11 other files it gives its agents.
- AGENTS.mdAGENTS.md
- asr3601-cross-branch-portingSkill
- asr3601-lvgl-firmware-triageSkill
- asr3601-project-onboardSkill
- asr3601-protocol-branch-matrixSkill
- asr3602-local-build-flashSkill
- asr360x-bug-delivery-orchestratorSkill
- catstudio-log-extractorSkill
- obsidian-fix-pattern-memorySkill
- skill-usage-trackerSkill
- zentao-bug-resolverSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Keep listing, deep investigation, and history reconciliation separate.
- Very specific: line 19
If the user already supplied full steps, actual/expected behavior, and local evidence, skip Zentao fetching and use …
- Distinctive: line 24
Run one command from the current firmware checkout:
- Distinctive: line 53
Use these only when explicitly requested:
- Very specific: line 78
Treat 已解决 as development-resolved and QA-pending.
- Distinctive: line 95
Read classification rules only for a deep classification, local storage only for cleanup or snapshot management, and project map only when …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Zentao Bug Triage
Keep listing, deep investigation, and history reconciliation separate. A plain list request must not preload every Bug's details or attachments.
Mode Selection
| Request | Mode | Work performed |
抓bug / 当前bug / 这个分支有哪些bug | List | Active rows and compact chat table only |
查 4045 / 修 4045 / selected IDs | Selected | Deep-fetch only those IDs and download their attachments |
全部深挖 / 回归对账 / historical status review | Deep/Reconcile | Explicit full details and exact-target reconciliation |
If the user already supplied full steps, actual/expected behavior, and local
evidence, skip Zentao fetching and use asr3601-lvgl-firmware-triage.
List Fast Path
Run one command from the current firmware checkout:
python -X utf8 "$env:USERPROFILE\.codex\skills\zentao-bug-triage\scripts\zentao_bug_fast_fetch.py" --repo .The default list path does not open Bug detail pages, download attachments,
search fix-pattern memory, create repair work queues, inspect code, or reconcile
history. Read chat-summary.md, paste its compact table, and stop so the user can
choose one Bug. Do not infer repair eligibility from a title-only row.
Selected Bug Path
Deep-fetch only the requested ID. Prefer one Bug at a time; accept multiple IDs only when the user explicitly selected them together.
python -X utf8 "$env:USERPROFILE\.codex\skills\zentao-bug-triage\scripts\zentao_bug_fast_fetch.py" --repo . --ids 4045This path opens the selected detail, records current status and activation
history, downloads its attachments, and creates work-items.md. Build one Bug
capsule containing title, status, steps, actual, expected, latest activation,
attachment paths, product, and observed time. Hand only that capsule to
asr3601-lvgl-firmware-triage; do not carry another Bug's attachments,
hypotheses, or repair eligibility into the task.
Deep And Reconcile Modes
Use these only when explicitly requested:
# Deep-fetch all listed Bugs and their attachments.
python -X utf8 "$env:USERPROFILE\.codex\skills\zentao-bug-triage\scripts\zentao_bug_fast_fetch.py" --repo . --deep-all
# Deep-fetch and compare the exact target with the previous local baseline.
python -X utf8 "$env:USERPROFILE\.codex\skills\zentao-bug-triage\scripts\zentao_bug_fast_fetch.py" --repo . --deep-all --reconcileReconciliation may refresh only previously tracked IDs missing from the current active list. Missing from a list is never itself a status transition.
Required Boundaries
- Load credentials only from the configured encrypted local credential. Never print or store passwords in reports, memory, Git, or this Skill.
- Resolve the current repository, branch, firmware identity, and exact Zentao product from live checkout evidence. Similar product-name prefixes are not an exact match.
- Fetch
activeby default. Fetch resolved, closed, all, or historical records only when requested. - List and investigation are read-only. They do not authorize source edits, branch changes, commits, device actions, releases, or Zentao writes.
- For a selected Bug, use the latest reactivation note as the current symptom.
Treat
已解决as development-resolved and QA-pending; only已关闭or explicit QA evidence supports QA verification. - Low-level hardware, modem, driver, or platform-owned evidence may be reported, but must not be converted into a firmware fix without a concrete owning path.
- Preserve unrelated local changes.
The launcher initializes missing project context, validates the resulting
snapshot against live Git and yl.h, and refreshes stale context only after a
successful fetch. Do not duplicate those checks outside the launcher.
Outputs And References
chat-summary.md: user-facing list; identify each item as<ID> <完整标题>.work-items.md: selected deep-fetch work order only.ignored-items.md: deep-mode skip/waiting reasons only.bugs.jsonandtriage.md: local snapshot evidence.
Read classification rules only for a deep classification, local storage only for cleanup or snapshot management, and project map only when the private current-project mapping is missing or ambiguous.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

