agentleFS
th3r-devth3r-dev/th3-skills/skills/qa/ticket-report/SKILL.md

qa-ticket-report

Skill2 starsChanged 2 days agoMITView on GitHub

Post a QA test or retest result onto a Jira ticket in the team format. The header block is Test Result / Test Environment (<env> - Version <actual>) / Platform / Test Date / Test Data / Actual Result. A ticket with more than one case gets a `Test Case | Actual Result | Expected Result` table. Under that the comment explains what happened, with the screenshots inline in the comment, not as an attachment list. Use WHENEVER the work is "post the test result on the ticket", "test report for PROJ-123", "Test Result: PASSED/FAILED", "retest done, log it on Jira", "รายงานผลเทสลงตั๋ว", "QA report standard", or after a qa/export-evidence or page-timing run whose result belongs on a ticket. Draft first, post only after the owner says OK.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Colour rule: every result word is bold and coloured, in the header and in each table row:

    line 36
  2. Very specific:

    Test Result takes the worst result among the judged cases, in the order FAILED > BLOCKED > NOT TESTED > PASSED.

    line 45
  3. Very specific:

    NOT TESTED > PASSED. In wiki markup that is {color:#00875A}*PASSED*{color}.

    line 46
  4. Very specific:

    More than one case = table: Write <dir>/cases.json from the ticket's acceptance criteria.

    line 55
  5. Very specific:

    Never put code paths, API names, commit hashes or column names in case / expected / actual / steps / why: they go in tech.

    line 71
  6. Very specific:

    Write it that way in actual.

    line 98

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

QA result comment on a ticket

S=skills/qa/ticket-report/scripts
uv run $S/ticket_report.py draft <evidence dir> --ticket PROJ-123 --env dev|uat|master|prd \
    --platform "Web (admin app)" --account "<section of ~/.config/qa-ticket-report/accounts.ini>" \
    --release vX.Y.Z --actual "<1-2 sentences>" [--lang th|en]
# show <dir>/ticket-report.md to the owner · on OK:
export JIRA_BASE_URL=https://<site>.atlassian.net JIRA_EMAIL=<you> JIRA_API_TOKEN=<token>
uv run $S/ticket_report.py post <evidence dir> --ticket PROJ-123 [--update <comment id>]
uv run $S/ticket_report.py --selftest

accounts.ini holds [Section] → UserName / Password and lives outside any repo. --test-data "<text>" prints a value as-is instead.

The format

references/format.md holds the template, a filled cases.json, and the shapes that come up.

  • The six header labels stay in English, exactly as written. The body follows the team's language (--lang th by default, Thai per the kien-thai rules; --lang en). No emoji anywhere.
  • Colour rule: every result word is bold and coloured, in the header and in each table row:

    ResultColourWhen
    PASSEDgreen #00875Athe case did what Expected says
    FAILEDred #DE350Bit ran and did something else
    BLOCKEDorange #FF8B00it could not run (env down, no data, no access)
    NOT TESTEDgrey #6B778Cin scope but not run this round

    Test Result takes the worst result among the judged cases, in the order FAILED > BLOCKED > NOT TESTED > PASSED. In wiki markup that is {color:#00875A}*PASSED*{color}.

  • Test Environment = <Dev|UAT|Master|PRD> - Version <actual>. "Actual" means what the tested app reports about itself: results.json version if present, else a read-only GET of the login page matched with --version-regex. Pre-release builds report a sha, so pass --release and the header becomes v1.2.22 (master-6ca0cf2). If the app says v1.2.21 and --release says something else, the draft stops. No version at all stops it too.
  • Test Data. On dev/uat it shows the username and the password. On master/prd the username only, and no flag overrides that. This is the one place a test password is printed; the drafts live in a gitignored work folder, and a leak check runs before any push.
  • More than one case = table. Write <dir>/cases.json from the ticket's acceptance criteria. evidence paths are relative to <dir> (../ is fine); every file has to exist. baseline: true marks a before-the-fix run kept for comparison: shown, not judged. With no cases.json, one case is built from results.json (page timing: every run 2xx within --max-seconds and the same row order) or from steps.csv (qa/export-evidence).
  • Write each case so a non-developer can read it in 30 seconds. Use the structured fields:

    fieldwhat goes inrule
    casewhat is tested, in plain wordsno codes, ids or field names
    expected / actualone sentence eachactual opens with the outcome
    stepslist, one action per itemwhat a tester clicks or enters, in order
    whyonly when not PASSED; 1–2 short sentencesthe cause in user terms
    datalist of name: valuedocument number, product, account, quantity
    techlist, for the dev teamcode paths, commit, API, table/column, ids

    Never put code paths, API names, commit hashes or column names in case / expected / actual / steps / why: they go in tech. One fact per sentence, no sentence over ~25 words.

  • Images go inline, each under its case with a one-line caption. Non-image files are listed under "attached files".
  • Actual Result under the header is one or two sentences with the numbers. Left out, the draft puts TODO there and post refuses to send.

How post works

  1. Uploads every evidence file as qa-<env>-<stamp>-c<case>-<name>; a name already attached is skipped, so a rerun is safe. Jira stores inline images as issue attachments.
  2. Sends POST /rest/api/2/issue/{key}/comment with a wiki-markup body (!name|width=700! for images, ||Test Case||…|| for the table). A markdown comment cannot embed images, which is why this goes through REST v2.
  3. Reads the comment back (expand=renderedBody) and prints inline images N/M. N < M means an image name did not match — check the upload names.

Traps

  • Every attachment must be previewable in Jira. Jira previews images, PDF, video and text/plain; application/octet-stream or application/json only downloads. preview_mime() sends .md .json .csv .log .txt .sql .tsv as text/plain. Never hard-code a MIME type.
  • Draft and post are separate steps on purpose. Nothing reaches Jira before the owner reads ticket-report.md.
  • A redraft keeps the upload names, so a re-post skips files already on Jira. To fix a posted comment in place use --update <comment id> rather than posting a second one.
  • A gateway 504 at exactly 60 s is the ingress cutting the request, not the API answering 504. Write it that way in actual.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.