create-observations
Skill1 starsChanged 2 months agoMITView on GitHub
Create or open today's observations file in dual-loop-brain/01-observations/ with the O-numbering continued from previous days. Use when the user wants to jot a new observation, idea, or piece of feedback about the app.
In this repositoryOne other file it gives its agents.
- write-handoffSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 21
If the file doesn't exist, find the highest O-number across the dated files (grep -h '^O[0-9]' …
- Very specific: line 26
If the user dictated observations in the invocation, write them in as O/C pairs — 2 lines each, unpolished.
- Distinctive: line 28
Keep those ids verbatim, even on a fresh file.
≈ “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
Step 0 — anchor: locate the repo root (the directory containing
dual-loop-brain/00-decision-log.md at or under cwd — 0 or 2+
matches → ask the human). If the session didn't start there, read
<root>/CLAUDE.md and <root>/dual-loop-brain/00-decision-log.md
explicitly — never assume the repo's CLAUDE.md loaded. The harness
guardrails arrive via the dual-loop plugin's session-start injection. Write only under
<root>/dual-loop-brain/01-observations/. Prefix shell commands with
cd <root> && — in some environments the cwd resets between commands
(never rely on a persistent cd). Inside the repo, its law
and decision log win over any workspace-level rules from the launch
directory.
/dual-loop:create-observations — today's observations file
date +%F→ today. Target:dual-loop-brain/01-observations/<today>-observations.md.- If the file doesn't exist, find the highest O-number across the dated
files (
grep -h '^O[0-9]' dual-loop-brain/01-observations/*-observations.md; the README's examples don't count; no dated files yet → start atO0, the app pitch — see the folder README). Create the file seeded with one empty pair:O<next>:/C<next>:. - If the user dictated observations in the invocation, write them in as O/C
pairs — 2 lines each, unpolished. Dictated with explicit ids (
O7: …)? Keep those ids verbatim, even on a fresh file. No ids → number onward from the next free one. Otherwise just report the file path and the next free id. - Nothing else: no beads, no backlog edits.
/dual-loop:design-sessiontakes it from here — it defaults to the newest file in the folder.
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.