note-context
Skill1 starsChanged 9 months agoMITView on GitHub
Save current session context to a SlashNote for handoff to next session
1 other project uses a version of this.
Adopted by
Most recently taken up by
- microsoft/cat-agent-skillstheir own version, first seen Nov 8, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- microsoft/cat-agent-skills★ 101same 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 repository10 other files it gives its agents.
- note-bugSkill
- note-decideSkill
- note-findSkill
- note-loopSkill
- note-meetingSkill
- noteSkill
- note-snippetSkill
- note-standupSkill
- note-todoSkill
- note-wrapupSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
14 rules. About 0 are common, 11 are distinctive, and 3 are 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.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 31
If found → save its id — will use mcp__slashnote__update_note in Step 3 instead of creating new
- Distinctive: line 32
If not found → will create a new note in Step 3
- Distinctive: line 43
Read note-loop note (if exists) → current tasks
- Distinctive: line 44
Read loop state file (if exists) → loop progress
- Very specific: line 49
If Step 0 found an existing context from current session (< 3 hours) → update it with mcp__slashnote__update_note
- Very specific: line 84
If no uncommitted changes → | clean
- Distinctive: line 87
If note-loop active → current task name + progress (N/M)
- Distinctive: line 88
If no loop → infer from conversation (what was being discussed/coded)
- Distinctive: line 108
Omit: if no decisions were made
- Distinctive: line 114
Omit: if no open questions
- Distinctive: line 134
If not in a git repo, skip git calls entirely:
- Distinctive: line 155
Always pin the note (context should stay visible)
- Distinctive: line 160
If user provides additional notes → add as "Notes" section before "Resume With"
- Distinctive: line 161
Keep the entire note scannable — this is a quick-reference card, not a detailed log
≈ “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
/note-context — Session Context Save
Save a snapshot of the current working context to a SlashNote. Designed for preserving state across Claude Code sessions — when context is about to be lost (compaction, session end, context switch).
Usage
/note-context # Auto-capture full context
/note-context <additional notes> # Auto-capture + user notesBehavior
Step 0: Find existing context to update
mcp__slashnote__search_noteswith query "Context"- Among results, find one where
previewstarts with "Context" ANDupdatedAtis less than 3 hours ago - If found → save its
id— will usemcp__slashnote__update_notein Step 3 instead of creating new - If not found → will create a new note in Step 3
Step 1: Collect context (one Bash call)
echo "PROJECT:$(basename $(git rev-parse --show-toplevel 2>/dev/null || pwd))" && echo "BRANCH:$(git rev-parse --abbrev-ref HEAD 2>/dev/null || echo 'no-git')" && echo "---STATUS---" && git status --short 2>/dev/null | head -20 && echo "---LOG---" && git log --oneline -5 2>/dev/nullStep 2: Collect SlashNote state
mcp__slashnote__list_notes— overview- Read note-loop note (if exists) → current tasks
- Read loop state file (if exists) → loop progress
Step 3: Create or update context note
Blue note, pinned:
- If Step 0 found an existing context from current session (< 3 hours) → update it with
mcp__slashnote__update_note - Otherwise → create new with
mcp__slashnote__create_note
# Context <timestamp>
**<project>** @ `<branch>` | <N> uncommitted changes
## Working On
<current task from note-loop, or inferred from conversation>
## Key Files
- `path/to/file.swift` — <what's being changed, 5 words max>
- `path/to/other.ts` — <purpose>
## State
- <branch purpose: "Adding X", "Fixing Y">
- <uncommitted changes: "N staged, M modified, K untracked">
- <loop progress: "/note-loop: N/M tasks done">
## Key Decisions
- <important decisions from this session>
## Open Questions
- <unresolved questions or uncertainties>
## Resume With
<exact command or action to pick up where we left off>Section Details
Header Line
- Format:
**<project>** @ \<branch>` |<N>uncommitted changes` - This gives instant orientation: what project, what branch, what's dirty
- If no uncommitted changes →
| clean
"Working On"
- If note-loop active → current task name + progress (N/M)
- If no loop → infer from conversation (what was being discussed/coded)
- Keep to 1-2 lines
- Be specific: "Improving /note-context skill in SlashNote plugin" not "Working on stuff"
"Key Files"
- NEW section — list the 3-5 most important files being worked on
- Source from
git status(modified files) and conversation context - Each file gets a brief annotation (max 5 words)
- This helps the next session immediately know where to look
- Skip if no specific files are relevant (e.g., research-only session)
"State"
- Branch purpose (infer from branch name:
feature/X→ "Adding X") - Uncommitted changes summary: "3 staged, 2 modified, 1 untracked"
- Loop progress if active
- Max 5 items
"Key Decisions"
- Decisions made in this session (check for decision notes)
- Technical choices, approach changes, rejected alternatives
- Omit if no decisions were made
"Open Questions"
- NEW section — unresolved questions or uncertainties
- Things that need clarification from the user or team
- Technical uncertainties that need investigation
- Omit if no open questions
"Resume With"
THE most critical section. Must be specific enough for a fresh Claude Code session to pick up immediately.
Quality bar: a new session reading ONLY this line should know exactly what to do next.
Good:
- "Continue editing
skills/note-loop/SKILL.md— adding note ID support" - "Run
/note-loop 3CEB5A5Bto continue skills roadmap" - "
git diffshows 3 modified files — review changes, then commit" - "Read
auth.swift:45-80, fix the JWT expiry check, then run tests"
Bad:
- "Continue working" (too vague)
- "Check the code" (meaningless)
- "Keep going" (useless)
Non-Git Mode
If not in a git repo, skip git calls entirely:
# Context <timestamp>
## Working On
<current task>
## Key Notes
- <relevant open notes summary>
## Open Questions
- <unresolved items>
## Resume With
<specific next action>Rules
- One Bash call max for git data
- Always pin the note (context should stay visible)
- Always blue color
- Timestamp includes time: "Feb 19, 2026 00:57"
- "Resume With" is mandatory — never skip it
- "Key Files" helps the next session navigate — always include if files are relevant
- If user provides additional notes → add as "Notes" section before "Resume With"
- Keep the entire note scannable — this is a quick-reference card, not a detailed log
- Optimized for LLM compaction — every line has maximum information density
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.

