surface-stale-tasks
Skill7 starsChanged 49 days agoMITView on GitHub
Flags to-dos from your task manager that may have gone stale (7+ days in your Today list, 21+ days anywhere else), using a full-detail export that includes each item's creation date. Read-only, conversational, never edits your task manager. Standalone, or run as a step from start-the-day. Trigger: /surface-stale-tasks, or asking to check for stale/forgotten tasks.
In this repository12 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- end-taskSkill
- end-the-daySkill
- link-taskSkill
- log-completed-tasksSkill
- pick-meetings-to-prepSkill
- prep-meetingSkill
- search-mailSkill
- setupSkill
- start-the-daySkill
- suggest-brag-itemsSkill
- update-tasksSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 51
Then stop and wait for an answer: Ask whether the user wants to move any forward, drop them, or leave them as-is, and end the turn there.
- Distinctive: line 58
If nothing is stale by these thresholds, just say so briefly rather than forcing the conversation.
- Distinctive: line 65
Never edit the task manager - no marking done, no rescheduling, no retagging.
- Distinctive: line 68
Never offer to make the change for them, even as a convenience - this skill only ever reads.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.
Files
Surface Stale Tasks
Purpose
Tasks can sit forgotten in a task manager - this is a nudge to notice, not a demand to act. It never edits your task manager itself; you decide what to do with anything flagged.
Deliberately uses a full-detail export rather than any human-readable
snapshot note you might also keep (see update-tasks) - a snapshot built for
readability may drop fields like creation date to stay scannable, but this
skill needs creation date to judge staleness. Don't substitute a trimmed-down
snapshot here even if it's faster to read; if it's missing the field this
skill depends on, the staleness check will be wrong.
Trigger
/surface-stale-tasks, or asking to check for stale/forgotten tasks. Also
invoked as a step by /start-the-day.
Process
- Pull two views from your task manager: your current "Today" list, and
every open item across all lists. Both should include each item's
creation date. Exactly how you get this out (AppleScript, an API call, a
CLI export) is specific to your tool -
skills/things3-export/has example scripts for Things3 on macOS. - From the Today list, flag anything whose creation date is 7+ days old - if it's been sitting in Today a week, it's probably not actually getting done there.
- From the full list, flag anything 21+ days old, excluding items already flagged from the Today list, so nothing is surfaced twice.
- Present these conversationally as "these look like they might have gone
stale" - not a demand. Two rules on how:
- Don't dump the full list as prose. The 21+ day list can easily run to dozens of items, and a wall of text gets skimmed past. Summarise it by area or project with counts, and name only the handful that genuinely stand out (a deadline about to bite, anything over a year old).
- Then stop and wait for an answer. Ask whether the user wants to
move any forward, drop them, or leave them as-is, and end the turn
there. Never ask and continue to the next step in the same message -
when called from
start-the-day, the caller must not move on to meetings until the user has actually replied. If they want to work through specific items, do that one at a time as structured questions rather than as another list.
- If nothing is stale by these thresholds, just say so briefly rather than forcing the conversation.
- Adjust the 7/21-day thresholds to taste - they're a starting point, not a law.
Guardrails
- Never edit the task manager - no marking done, no rescheduling, no retagging. That's the user's action to take, in the tool that owns that state.
- Never offer to make the change for them, even as a convenience - this skill only ever reads.
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
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.

