systems-review
Skill0 starsChanged 7 months agoMITView on GitHub
Weekly review of all active systems — GTD, Atomic, PQ, Friction Elimination, Intuitive Eating. Walk through each system spec, check adherence, capture learnings, and update specs.
In this repository9 other files it gives its agents.
- obsidian-gtd-cliAGENTS.md
- clarifySkill
- nextSkill
- organizeSkill
- projectSkill
- purgeSkill
- reviewSkill
- stuckSkill
- task-prioritisationSkill
The file
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 11 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈11
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
One question at a time: Never combine two questions in one message.
- Distinctive: line 21
Use concrete, low-friction questions: Ask "Did X happen?" or "How many days did you do Y?" — not "What do you want to do more/less of?" If …
- Distinctive: line 22
Capture notes proactively: Don't wait for the user to remind you — take status notes after each system and show them before moving on.
- Distinctive: line 23
When the user mentions a concept they can't remember,: search the web before answering from memory.
- Very specific: line 28
Read the current state of all system specs using the /obsidian skill:
- Distinctive: line 49
Keep it conversational. Don't interrogate — just ask one question at a time and follow the thread.
- Distinctive: line 67
Run the resistance map live if needed.
- Distinctive: line 74
After discussing each system, append a dated entry to the Review Log section of that system's spec:
- Distinctive: line 83
If the user wants to change a routine, tweak a cue, add or remove something — update the spec directly.
- Distinctive: line 94
Don't force all five if the user only has energy for a few — prioritise what needs attention
- Distinctive: line 95
If a system is running smoothly, a quick "GTD's solid this week, anything to flag?" is enough
- Distinctive: line 97
Keep review log entries short and actionable, not essay-length
≈ “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
Systems Review
Interactive weekly review of all active systems. Run on Sundays during the 16:00–17:00 review block.
Overview
The systems index lives at References/Systems.md in the vault. Each system has its own spec in References/ with routines, principles, and a review log.
How to Run
This is a conversational skill. Walk through each system one at a time with the user. Don't dump everything at once — ask questions, listen, and capture.
Conversation Rules
- One question at a time. Never combine two questions in one message.
- Use concrete, low-friction questions. Ask "Did X happen?" or "How many days did you do Y?" — not "What do you want to do more/less of?" If the user stalls or says "too much friction", rephrase with less abstraction.
- Capture notes proactively. Don't wait for the user to remind you — take status notes after each system and show them before moving on.
- When the user mentions a concept they can't remember, search the web before answering from memory.
- Keep it moving. The review itself is a friction source. If a system is running well, one question is enough — don't force depth where there's nothing to discuss.
Step 1 — Load the specs
Read the current state of all system specs using the /obsidian skill:
# See /obsidian skill for $OBS and $VAULT setup
$OBS vault=$VAULT read path="References/Systems.md"
$OBS vault=$VAULT read path="References/GTD System.md"
$OBS vault=$VAULT read path="References/Atomic System.md"
$OBS vault=$VAULT read path="References/PQ System.md"
$OBS vault=$VAULT read path="References/Friction Elimination.md"
$OBS vault=$VAULT read path="References/Intuitive Eating System.md"Step 2 — Walk through each system
For each system, ask:
- How did it go this week? — Did you follow the routines?
- What worked? — Anything to keep or reinforce?
- What didn't work? — What got missed and why?
- What needs to change? — Any tweaks to the spec?
Keep it conversational. Don't interrogate — just ask one question at a time and follow the thread.
System order
- GTD — Start here. Check metrics using the
/obsidianskill's Dataview queries (inbox count, overdue, stale projects, context distribution, completed this week). Key questions: inbox at zero? Projects moving? Any stale? Overdue tasks? - Atomic — Morning and evening routines. Key questions: how many days did you do the morning routine? Evening routine? What broke the chain?
- PQ — Mindfulness and refreshes. Key questions: did the 08:00/13:00/16:00 happen? Any saboteurs you noticed this week?
- Friction Elimination — Resistance mapping.
Key questions (use these concrete prompts):
- "Is there a task you keep seeing and skipping?"
- "Did anything break or get annoying this week?"
- "Did something just work this week without effort?" Run the resistance map live if needed. Log any actions taken.
- Intuitive Eating — Weekly check-in. Key questions: use the prompts in the spec. How's the relationship with food? Diet mentality creeping in?
Step 3 — Capture to review logs
After discussing each system, append a dated entry to the Review Log section of that system's spec:
### YYYY-MM-DD
- [what was discussed, what changed, what to try next week]Step 4 — Update specs if needed
If the user wants to change a routine, tweak a cue, add or remove something — update the spec directly. The specs are living documents.
Step 5 — Summary
At the end, give a brief summary:
- Which systems are running well
- Which need attention next week
- Any spec changes made
Tips
- Don't force all five if the user only has energy for a few — prioritise what needs attention
- If a system is running smoothly, a quick "GTD's solid this week, anything to flag?" is enough
- The Friction Elimination review can double as a live coaching session — help the user break down stuck tasks on the spot
- Keep review log entries short and actionable, not essay-length
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.

