Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Track technical changes with structured records, a state machine, and session handoff. Usage: /tc <init|create|update|status|resume|close|export|dashboard> [args]
.claude/skills/alirezarezvani-tc/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-18 | ✗→✓ | ▲ Improved | 63% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 83% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 104% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 14% | 0% |
Dispatch a TC (Technical Change) command. Arguments: $ARGUMENTS.
If $ARGUMENTS is empty, print this menu and stop:
/tc init Initialize TC tracking in this project
/tc create <name> Create a new TC record
/tc update <tc-id> [...] Update fields, status, files, handoff
/tc status [tc-id] Show one TC or the registry summary
/tc resume <tc-id> Resume a TC from a previous session
/tc close <tc-id> Transition a TC to deployed
/tc export Re-render derived artifacts
/tc dashboard Re-render the registry summaryOtherwise, parse $ARGUMENTS as <subcommand> <rest> and dispatch to the matching protocol below. All scripts live at engineering/tc-tracker/scripts/.
initbash python3 engineering/skills/tc-tracker/scripts/tc_init.py --root . --json
already_initialized, report current statistics and stop./tc create <name> as the next step.create <name><name> as a kebab-case slug. If missing, ask the user for one.feature | bugfix | refactor | infrastructure | documentation | hotfix | enhancementcritical | high | medium | low (default medium)bash python3 engineering/skills/tc-tracker/scripts/tc_create.py --root . \ --name "<slug>" --title "<title>" --scope <scope> --priority <priority> \ --summary "<summary>" --motivation "<motivation>" --json
update <tc-id> [intent]<tc-id> is missing, list active TCs (status in_progress or blocked) from tc_status.py --all and ask which one.--set-status <state> with --reason "<why>"--add-file path[:action]--add-test "<title>" --test-procedure "<step>" --test-expected "<result>"--handoff-progress, --handoff-next, --handoff-blocker, --handoff-context--note "<text>"--tag <tag>bash python3 engineering/skills/tc-tracker/scripts/tc_update.py --root . --tc-id <tc-id> [flags] --json
status [tc-id]<tc-id> is provided:bash python3 engineering/skills/tc-tracker/scripts/tc_status.py --root . --tc-id <tc-id>
bash python3 engineering/skills/tc-tracker/scripts/tc_status.py --root . --all
resume <tc-id>bash python3 engineering/skills/tc-tracker/scripts/tc_status.py --root . --tc-id <tc-id> --json
progress_summary, next_steps (numbered), blockers, key_context.bash python3 engineering/skills/tc-tracker/scripts/tc_update.py --root . --tc-id <tc-id> \ --note "Session resumed" --reason "session handoff"
next_steps. Do NOT re-derive context — trust the handoff.close <tc-id>tc_status.py --tc-id <tc-id> --json.tested. If not, refuse and tell the user which transitions are still required.test_cases: warn if any are pending, fail, or blocked.bash python3 engineering/skills/tc-tracker/scripts/tc_update.py --root . --tc-id <tc-id> \ --set-status deployed --reason "Approved by <approver>" --note "Approval: <approver> — <notes>" Then directly edit the approval block via a follow-up update if your script version supports it; otherwise instruct the user to record approval in notes.
exportThere is no automatic HTML export in this skill. Re-validate everything instead:
bash python3 engineering/skills/tc-tracker/scripts/tc_validator.py --record <path> --json
bash python3 engineering/skills/tc-tracker/scripts/tc_validator.py --registry docs/TC/tc_registry.json --json
dashboardRun the all-records summary:
bashpython3 engineering/skills/tc-tracker/scripts/tc_status.py --root . --all
tc_record.json by hand. Always use tc_update.py so revision history is appended and validation runs.[CANCELLED].engineering/tc-tracker — Full SKILL.md with schema reference, lifecycle diagrams, and the handoff format.engineering/changelog-generator — Pair with TC tracker: TCs for the per-change audit trail, changelog for user-facing release notes.engineering/tech-debt-tracker — For tracking long-lived debt rather than discrete code changes.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | pass→pass | 10,756 | 8,711 | -19% | 1 | 1 | 0% | 2,042 | 3,187 | +56% | 0 | 0 | — |
case-18 | fail→pass | 8,519 | 2,909 | -66% | 1 | 1 | 0% | 1,318 | 2,149 | +63% | 0 | 0 | — |
case-01 | fail→fail | 11,101 | 5,493 | -51% | 1 | 1 | 0% | 1,743 | 2,640 | +51% | 0 | 0 | — |
case-02 | pass→pass | 14,014 | 4,398 | -69% | 1 | 1 | 0% | 2,283 | 2,422 | +6% | 0 | 0 | — |
case-04 | fail→fail | 3,052 | 4,841 | +59% | 1 | 1 | 0% | 582 | 2,037 | +250% | 0 | 0 | — |
case-05 | fail→fail | 7,541 | 6,197 | -18% | 1 | 1 | 0% | 1,165 | 2,111 | +81% | 0 | 0 | — |
case-06 | fail→fail | 7,724 | 4,290 | -44% | 1 | 1 | 0% | 1,461 | 1,897 | +30% | 0 | 0 | — |
case-07 | fail→fail | 6,470 | 1,907 | -71% | 1 | 1 | 0% | 1,121 | 2,069 | +85% | 0 | 0 | — |
case-19 | fail→pass | 9,097 | 3,320 | -64% | 1 | 1 | 0% | 1,522 | 2,288 | +50% | 0 | 0 | — |
case-08 | fail→fail | 10,221 | 3,379 | -67% | 1 | 1 | 0% | 1,766 | 2,335 | +32% | 0 | 0 | — |
case-09 | fail→pass | 6,340 | 2,771 | -56% | 1 | 1 | 0% | 1,246 | 2,283 | +83% | 0 | 0 | — |
case-10 | fail→pass | 7,218 | 4,851 | -33% | 1 | 1 | 0% | 1,329 | 2,712 | +104% | 0 | 0 | — |
case-11 | fail→fail | 8,550 | 4,317 | -50% | 1 | 1 | 0% | 1,522 | 2,608 | +71% | 0 | 0 | — |
case-12 | fail→pass | 10,269 | 2,192 | -79% | 1 | 1 | 0% | 1,785 | 2,040 | +14% | 0 | 0 | — |
case-13 | pass→pass | 5,518 | 2,673 | -52% | 1 | 1 | 0% | 971 | 2,190 | +126% | 0 | 0 | — |
case-14 | pass→pass | 9,576 | 3,551 | -63% | 1 | 1 | 0% | 1,661 | 2,354 | +42% | 0 | 0 | — |
case-15 | fail→fail | 8,948 | 5,018 | -44% | 1 | 1 | 0% | 1,670 | 2,642 | +58% | 0 | 0 | — |
case-16 | fail→pass | 11,532 | 4,234 | -63% | 1 | 1 | 0% | 2,048 | 2,451 | +20% | 0 | 0 | — |
case-17 | fail→pass | 14,536 | 1,272 | -91% | 1 | 1 | 0% | 2,571 | 1,847 | -28% | 0 | 0 | — |
case-20 | fail→pass | 8,007 | 2,359 | -71% | 1 | 1 | 0% | 1,342 | 2,118 | +58% | 0 | 0 | — |
case-21 | fail→pass | 11,675 | 2,343 | -80% | 1 | 1 | 0% | 2,000 | 2,156 | +8% | 0 | 0 | — |
case-22 | pass→pass | 6,393 | 3,489 | -45% | 1 | 1 | 0% | 1,029 | 2,374 | +131% | 0 | 0 | — |
case-23 | pass→pass | 7,093 | 2,138 | -70% | 1 | 1 | 0% | 1,359 | 2,133 | +57% | 0 | 0 | — |
case-24 | pass→pass | 8,017 | 2,058 | -74% | 1 | 1 | 0% | 1,291 | 2,151 | +67% | 0 | 0 | — |
case-25 | fail→fail | 7,511 | 2,372 | -68% | 1 | 1 | 0% | 1,240 | 2,063 | +66% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 25 cases were attempted, and 22 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.