save-session-log
Skill2 starsChanged 8 days agoMITView on GitHub
Keep conversation records and decision traces. Use when setting up transcript hooks, when the user asks where a decision came from, or when a host has no hook support.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- TeleginS/agent-orchestrationtheir own version, first seen Nov 16, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- TeleginS/agent-orchestration★ 6same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository4 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- maintain-project-stateSkill
- spec-workflowSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
Run python .agents/tools/ws.py hooks enable <host> once per tool.
- Distinctive: line 18
Do not retype the conversation.
- Very specific: line 35
After transcripts are copied, ws reconcile lists each user message that no row accounts for.
- Very specific: line 36
Read each one. If it holds a requirement, decision or preference that is not recorded, record it first.
- Distinctive: line 38
Do this before the final checkpoint of a session.
- Very specific: line 42
Never store hidden reasoning or private host data.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
Save session log
Conversations are records. Decisions from them belong in the state tables.
With hooks
Run python .agents/tools/ws.py hooks enable <host> once per tool.
The hooks copy each transcript to docs/chats/<host>/<session>.md and list it in docs/chats/INDEX.md.
Nothing else is needed during the session.
Without hooks
Do not retype the conversation. Record its results instead:
| From the conversation | Record with |
| A decision | ws add decision --set source="user 2026-10-31: <short quote>" |
| A requirement change | ws add or ws set with --source "user 2026-10-31" |
| An open question | ws add question |
| A preference | ws add preference |
The source field keeps a short original quote in the user's language. That is enough to trace a fact back to the conversation. A credential in the quote follows the project's credential rule.
ws start imports Kimi, Codex and Kiro sessions on its own. Cursor and Antigravity sessions are
imported with ws chats import cursor or ws chats import antigravity.
Reconcile
After transcripts are copied, ws reconcile lists each user message that no row accounts for.
Read each one. If it holds a requirement, decision or preference that is not recorded, record it
first. Then mark it with the IDs, or with a reason such as chat: thanks or question: status.
Do this before the final checkpoint of a session.
Rules
- Never store hidden reasoning or private host data. The capture redacts common English and Chinese forms
of credentials, such as
password: ...and密码:..., not every secret. It does so also under"credentials": "plain", which keeps credentials as written only in the files the user named. - A transcript does not override the state files.
- Raw copies in
docs/chats/*/raw/stay out of version control.
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.

