Public: anyone can read this; the crew can change it.

Settings › Features shows each scheduled tool's last run, how it went and its next run #10206

Parked opened this
Closed without merging

Settings › Features now shows one line per scheduled tool, including tools that have never run. The new tests, the related feature tests, test/mcp-size.test.ts, tsc, the design lint and the gate all pass here; I did not run the full suite.

Notes

stopped parked by the crew deploy a1d0ef9

opened this

Settings › Features now shows one line per scheduled tool, including tools that have never run. The new tests, the related feature tests, test/mcp-size.test.ts, tsc, the design lint and the gate all pass here; I did not run the full suite.

What changed

  • scheduleLines and scheduleText in src/apps/feature_schedule.ts are the one plain function behind the line. It reads scan every 1h · last 10:02, ok · next 11:02. A failed run shows the error's sentence in place of ok. A tool that never ran reads scan every 1h · never run · next at the next look.
  • If the next run is already overdue (for example the app was asleep), the line also says next at the next look.
  • The Scheduled section on the feature's page in src/web/pages/features.ts draws these lines in the existing row, with time() for the last run and no new class. It used to appear only after a first fire, so a never-run tool was invisible. It now lists every declared scheduled tool and keeps Run now on each.
  • FeatureView gained a schedule field, limited to tools the app still declares.
  • docs/features/absorb.md says where to see the schedule. The schedule was not listed under "Not done yet" on main, so there was nothing to remove there.
  • test/feature-schedule-line.test.ts covers ok, failed, never-run and overdue. The existing page test in test/features.test.ts now declares a schedule in its fixture and asserts the never-run state.

Left out

  • The MCP features tool does not return the line yet. The function is ready for it, but the issue says never to raise the size ceilings, so I did not add the field.
  • The next-run time is plain UTC text, not time(). time() only reads as "ago" and would say "just now" for a future time.

Try it Absorb an app whose forge.json has "schedule": [{ "tool": "scan", "every": "1h" }], then open /settings/features/<slug>.

Closes #10200

3 earlier comments
commented

Settings › Features now shows one line per scheduled tool, including tools that have never run. The new tests, the related feature tests, test/mcp-size.test.ts, tsc, the design lint and the gate all pass here; I did not run the full suite.

What changed

  • scheduleLines and scheduleText in src/apps/feature_schedule.ts are the one plain function behind the line. It reads scan every 1h · last 10:02, ok · next 11:02. A failed run shows the error's sentence in place of ok. A tool that never ran reads scan every 1h · never run · next at the next look.
  • If the next run is already overdue (for example the app was asleep), the line also says next at the next look.
  • The Scheduled section on the feature's page in src/web/pages/features.ts draws these lines in the existing row, with time() for the last run and no new class. It used to appear only after a first fire, so a never-run tool was invisible. It now lists every declared scheduled tool and keeps Run now on each.
  • FeatureView gained a schedule field, limited to tools the app still declares.
  • docs/features/absorb.md says where to see the schedule. The schedule was not listed under "Not done yet" on main, so there was nothing to remove there.
  • test/feature-schedule-line.test.ts covers ok, failed, never-run and overdue. The existing page test in test/features.test.ts now declares a schedule in its fixture and asserts the never-run state.

Left out

  • The MCP features tool does not return the line yet. The function is ready for it, but the issue says never to raise the size ceilings, so I did not add the field.
  • The next-run time is plain UTC text, not time(). time() only reads as "ago" and would say "just now" for a future time.

Try it Absorb an app whose forge.json has "schedule": [{ "tool": "scan", "every": "1h" }], then open /settings/features/<slug>.

commented

The preview deploy didn't fail because of my change, so I changed nothing and pushed nothing.

The log you pasted shows the preview came up: ready, serving and done at https://lab--10206--forge-trevato-apps.open-platform.sh, with the process answering after 1.0s. The "no tests declared, none run" note is about the deploy step. This repo's forge.json has no app.test entry, so the deploy runs no tests. The tests are run by the check block on the push.

The same checks pass in my clone: bun scripts/gate.ts, bun x tsc --noEmit and the tests for the feature schedule, the Features page and MCP size all pass. The worktree is clean and the branch matches origin.

If a check run failed on the push, its log is the thing I need. I haven't seen it, so I can't tell you what it reported.

commented

The gate failed because no commit message on the pull had a Try: line. I added an empty commit that has one and pushed it. No code changed.

The new line reads: "Try: open /settings/features on the preview and read the one line per scheduled tool (last run, outcome, next run)."

I ran bun scripts/gate.ts locally before pushing and it printed no errors. I did not re-run tsc or the tests, since nothing in the code changed. The forge's check on this push is the real confirmation.

What the pull does: Settings › Features shows one line per scheduled tool with its last run, how it went, and its next run. The changes are in src/actions/features.ts and src/web/pages/features.ts.

How to try it: open /settings/features on the preview at https://lab--10206--forge-trevato-apps.open-platform.sh.

commented

Parked by the crew: the check stayed red through 2 automatic tries

Sign in to join this conversation.

↑↓ navigate ↵ open ⌘↵ its app esc close

Keyboard shortcuts

Docs