CtrlK
BlogDocsLog inGet started
Tessl Logo

testland/bug-tracker-workflow

Files, transitions, and searches bugs in the team's tracker - Jira, Linear, GitHub Issues, or Azure DevOps - through one tracker-agnostic workflow: authenticate, dedupe-search before creating, create with severity/priority classification, transition lifecycle states, and wire idempotent CI-driven filing from test failures. Jira Cloud REST API v3 is worked in full in the body (ADF descriptions, runtime transition lookup, JQL triage and duplicate queries, dry-run bulk transitions); Linear's GraphQL API (issueCreate/issueUpdate, workflowStates resolved by type, the 0-4 priority enum), GitHub Issues REST (open/closed + state_reason, label-based severity/priority), and Azure DevOps Work Item Tracking (JSON Patch, WIQL, process-template states) each have a deep reference. Use when programmatically managing the bug lifecycle on any of the four trackers: creating from CI failures, triaging queues, transitioning states, or dedupe-searching.

94

1.01x
Quality

96%

Does it follow best practices?

Impact

94%

1.01x

Average score across 10 eval scenarios

SecuritybySnyk

Low

Low-risk findings worth noting

Overview
Quality
Evals
Security
Files

task.mdevals/scenario-7/

A token-leak ticket has been sitting in the cosmetic pile for five weeks

Problem Description

While preparing for a customer security questionnaire I went through the open backlog and found a ticket about password-reset tokens ending up in third-party analytics filed under our cosmetic bucket, priority Low, tagged as a UI issue. It has been open 37 days.

That prompted a wider look. Our tracker has a built-in Priority field and a separate Severity field, and the board is a mess: some tickets have one filled and not the other, at least one has both stuffed into the title, and there is a long-running argument on the team that the two fields should always agree, which I do not think is right - we deliberately do not drop everything for a crash that affects two percent of one platform.

I need a classification pass I can defend to the security reviewer and to the engineers whose tickets I am about to change.

Output Specification

Produce exactly two files:

  1. classification-review.md - per ticket that needs a change: what is wrong with its current classification, what it should be, and the basis for each value separately. Address the "the two fields should always agree" argument directly, using tickets from this export as evidence. List the tickets whose current values are correct as they stand, including any where the two fields deliberately diverge, and say why they must not be adjusted.
  2. reclassification.csv - one row per field being changed, columns id,field,current_value,proposed_value,basis. A ticket with two fields changing produces two rows.

Out of scope: assignment, scheduling, remediation design, and any state transition. Do not propose closing anything.

Input Files

Extract the following files before beginning.

=============== FILE: exports/open-backlog.csv =============== id,title,labels,priority_field,severity_field,state,assignee,age_days,evidence WEB-512,"Password reset link is placed in the page URL and forwarded in the Referer header","ui,cosmetic",Low,5 - Cosmetic,Open,,37,"reset tokens found in the analytics vendor's export; tokens valid 30 minutes; any vendor staff with report access can complete a reset" WEB-519,"Session cookie loses the Secure attribute on the staging-to-prod redirect","backend",High,,Open,d.novak,12,"cookie observed over plain http on the redirect hop; reproduced on 4.18.2" WEB-523,"Settings page shows 'Sav e' instead of 'Save'","ui,cosmetic",Low,5 - Cosmetic,Open,,20,"screenshot; text only, no functional effect" WEB-528,"Data export returns rows belonging to other tenants when the filter is cleared","backend,data",Highest,,Assigned,s.park,4,"confirmed on staging with two seeded tenants; 1,400 rows from tenant B in tenant A's export" WEB-531,"Checkout unavailable for all users in the EU region during peak hours","backend",Low,1 - Critical,Open,m.oyelaran,6,"90 minutes of total outage on the eu cluster 2026-08-11; downgraded to Low because it was reported by an internal user rather than a customer" WEB-535,"App crashes on launch for Android 9 devices","mobile",Medium,1 - Critical,Assigned,f.okafor,15,"hard crash, no workaround, affects 2.1% of installs; Android 9 is 1.4% of revenue and drops off the support matrix in November" WEB-540,"[P1][CRITICAL] Audit log missing entries for permission changes","backend",,,Open,,9,"3 permission changes on 2026-08-06 absent from the log; required for the SOC 2 evidence pack" WEB-546,"Tooltip arrow misaligned by 2px on the billing page","ui,cosmetic",Low,5 - Cosmetic,Open,,44,"screenshot"

SKILL.md

tile.json