Tracking Fullstack Worker

What it does

When your agent uses it

• Tasks that involve Time tracking and reporting

Example prompts

• “/tracking-fullstack-worker”

Workflow steps

12 steps, taken from the first numbered list in SKILL.md.

1. 1Read mission mission.md, mission AGENTS.md, .factory/services.yaml, .factory/library/architecture.md, .factory/library/user-testing.md…
2. 2Convert the feature's expectedBehavior bullets into one shared seeded scenario that spans backend truth and browser-visible behavior.
3. 3Read only the relevant source docs from the closed set in mission AGENTS.md.
4. 4Write the failing tests first on both sides as needed
5. 5Reuse the running backend on 8080 and website on 5173; do not start replacement main runtimes.
6. 6Implement only after the failing tests prove the intended gap.
7. 7Keep the same seeded fact across layers. Do not prove the backend and browser with unrelated scenarios.
8. 8Unless the feature description explicitly says it is proof-only, update the status blocks for every directly used source doc relied on in…
9. 9For every backend+browser feature, keep both halves of the proof. Do not hand off a browser-only result for a fullstack feature unless the…
10. 10Run both the backend and frontend verification commands touched by the feature.
11. 11Use agent-browser for one end-to-end replay of the shared scenario after automated tests pass.
12. 12In the handoff, explicitly call out

What it can do on your machine

Read from SKILL.md and the folder at commit 5ffd0a0. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md (its code samples are json).

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Tracking Fullstack Worker loads about 1.3k tokens when it runs. Until then it costs about 34 tokens; SKILL.md has 377 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~34

When it runs · the whole SKILL.md, loaded when a task matches

~1.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from CorrectRoadH/OpenTickly at commit 5ffd0a0, republished under its AGPL-3.0 licence (© CorrectRoadH). 377 words, ~1,329 tokens.

name

tracking-fullstack-worker

description

Implement timer/time-entry behavior that must be proven across backend truth and the browser surface together.

{
  "salientSummary": "Implemented the cross-workspace running-header flow so a timer started in workspace A stays visible and operable after switching to workspace B while timer history re-scopes to workspace B only. Added one backend truth proof and one browser replay built on the same seeded two-workspace scenario.",
  "whatWasImplemented": "Changed the timer experience so the global running timer remains visible, editable, and stoppable across workspace switches while calendar/list/timesheet history stays filtered to the current workspace. The work paired backend regressions for the current-timer fact with real-runtime browser coverage for the top composer and history projections, and updated the relevant tracking/timer source-doc status blocks.",
  "whatWasLeftUndone": "",
  "verification": {
    "commandsRun": [
      {
        "command": "go test ./apps/backend/internal/tracking/... -count=1 -parallel 4",
        "exitCode": 0,
        "observation": "Backend current-timer and stop/readback proofs passed."
      },
      {
        "command": "vp run test:e2e:website -- e2e/timer-page.spec.ts --workers 1",
        "exitCode": 0,
        "observation": "Browser proof for cross-workspace running header and workspace-scoped history passed."
      },
      {
        "command": "vp run check -r",
        "exitCode": 0,
        "observation": "Repo JS/TS checks remained green after the fullstack change."
      }
    ],
    "interactiveChecks": [
      {
        "action": "Started a timer in workspace A, switched to workspace B, edited and stopped the timer from the shared header, and observed the history views.",
        "observed": "The running timer stayed visible and operable across the switch, and the history projection remained scoped to workspace B without leaking the stopped workspace-A entry."
      }
    ]
  },
  "tests": {
    "added": [
      {
        "file": "apps/backend/internal/tracking/application/regressions_test.go + apps/website/e2e/timer-page.spec.ts",
        "cases": [
          {
            "name": "workspace switch preserves the global running header while history re-scopes",
            "verifies": "VAL-CROSS-001"
          },
          {
            "name": "stopping a foreign-workspace running timer clears current state without leaking history",
            "verifies": "VAL-CROSS-003"
          }
        ]
      }
    ]
  },
  "discoveredIssues": []
}

© CorrectRoadH, AGPL-3.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .factory/skills/tracking-fullstack-worker of CorrectRoadH/OpenTickly.

Open the folder on GitHubat commit 5ffd0a0