touch
Skill1 starsChanged 19 days agoApache-2.0View on GitHub
Agent-callable single-fact update to a standing file right after the work it records; one queue status, flag or state line, plus a ledger line. Never moves, resolves, rewrites or runs checks. Does not replace or satisfy session-close.
In this repository21 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- .cursor/rules/agents.mdcCursor rule
- commandsSkill
- daily-dashboardSkill
- daily-reviewSkill
- document-standardsSkill
- jotSkill
- merge-proxySkill
- monthly-reviewSkill
- orientSkill
- redaction-gateSkill
- session-closeSkill
- session-logSkill
- skill-creationSkill
- skill-discoverySkill
- surface-regimesSkill
- technical-documentsSkill
- weekly-reviewSkill
- working-preferencesSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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 41
Run the dash sweep or any other check, or state that one passed.
≈ “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
When to Use
Immediately after the work a standing file should now reflect: a task finished and verified by a
second party, an open question raised, a state change a later session must see. Called by the agent.
The shared rules, marker and placement are in docs/partial-captures.md; read that first.
The allowed edits
One fact per call, and only these kinds:
| Standing file | Allowed |
| Task queue | Set one task's status, with a dated note saying what verified it and who the second party was |
| Open-threads log | Append one dated group before the resolved section |
| Manifest | Append one line to the current state section |
The fact must have come from this session's work. A touch records what just happened; it does
not correct what an earlier session wrote. A wrong line found in a standing file goes into the
open-threads log as a finding, and session-close fixes it.
Then append to the day's capture file:
## HH:MM [touch] <file>: <the fact, in a few words>Time from date. Build every edit with a script that asserts its anchor occurs exactly once, so a
missing or doubled anchor fails loudly instead of writing in the wrong place.
What it must not do
- Mark a task complete on the claimant's word alone. Complete needs a second party, named in the note.
- Move a thread to the resolved log, delete a flag, or rewrite an existing line. Those are judgment
over the whole record, and they are
session-close's. - Run the dash sweep or any other check, or state that one passed.
- Touch the operating handoff document. Its roster and conventions change by decision, not by event.
Scope Pointer
session-close: the full update of every standing file, with the sweep and the checks. It resets the partial count, and a stack of touches does not discharge it.jot: the other partial. It records reasoning; this records facts.daily-dashboard: reads the standing files and checks them against the tree, so a touch that got a fact wrong is caught there if not before.
<!--
SPDX-FileCopyrightText: 2026 Aidan Naveja
SPDX-License-Identifier: Apache-2.0
-->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.