session-context
Skill4 starsChanged 7 months agoMITView on GitHub
Load persistent codebase memory at session start. Automatically recalls project conventions, architecture decisions, and context from previous sessions. When an agent needs to: (1) Start a new session with full project context, (2) Recall what was learned in prior sessions, (3) Restore project memory, or (4) Initialize Tages for a new project. Gracefully offers setup if Tages isn't configured.
7 other projects use a version of this.
Adopted by
Most recently taken up by
- mahumadad/quarrytheir own version, first seen Nov 8, 2026
- michaelalber/ai-toolkittheir own version, first seen Nov 1, 2026
- Vinay-R/vibe-engineering-templatetheir own version, first seen Nov 1, 2026
- pskoett/pskoett-ai-skillstheir own version, first seen Oct 25, 2026
- sandraschi/onenote-mcptheir own version, first seen Oct 9, 2026
- InsightStudio-2026/agent-assetstheir own version, first seen Jul 1, 2026
Other versions7 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 7
- pskoett/pskoett-ai-skills★ 327same name
- paolomoz/skills★ 4same name
- michaelalber/ai-toolkit★ 3same name
- InsightStudio-2026/agent-assets★ 1same name
- mahumadad/quarry★ 1same name
- sandraschi/onenote-mcp★ 1same name
- Vinay-R/vibe-engineering-template★ 0same 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 repository2 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- proactive-memorySkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
3
Distinctive
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 23
Call the brief tool to load a ranked, token-budgeted project context:
- Distinctive: line 39
Do not dump the full memory contents unless asked.
- Very specific: line 43
If the tools fail because no Tages configuration exists for this project, offer setup:
- Distinctive: line 52
If the user says yes, run the appropriate init command.
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Session Context — Restore Codebase Memory
Keywords
session start, project context, recall, what do you know about this project, load context, restore memory, previous session, remember, what did we learn, project conventions, codebase memory
Overview
Restore persistent codebase memory at the start of a session. Calls Tages MCP tools to hydrate the agent with conventions, architecture decisions, and project context from previous sessions. If Tages isn't configured for the current project, offers a smooth setup flow.
Use this skill when: Starting a new session, or when the user asks what you know about the project.
Workflow
Step 1: Hydrate Context
Call the brief tool to load a ranked, token-budgeted project context:
brief()This returns the most important memories — gotchas, conventions, architecture decisions, and patterns — ranked by confidence, recency, and usage. High-value memories appear in full detail; lower-priority ones are summarized as compact key-value pairs. The token budget adapts to project size automatically.
Step 2: Handle the Result
Tool returns data (happy path):
Absorb the returned brief into your working context. Summarize briefly to the user:
Loaded project context — conventions, architecture decisions, and prior knowledge.
Do not dump the full memory contents unless asked. Use them silently to inform your work.
Tools error — Tages not configured (setup path):
If the tools fail because no Tages configuration exists for this project, offer setup:
Tages isn't configured for this project yet. I can set it up so I'll remember conventions, architecture decisions, and context across sessions.
- Cloud (recommended):
npx @tages/cli init— syncs to dashboard, works across machines- Local only:
npx @tages/cli init --local— SQLite only, no account neededWant me to set it up?
If the user says yes, run the appropriate init command. If no, proceed without memory and do not ask again this session.
Tools succeed but return empty:
No memories stored yet for this project. I'll start remembering conventions and decisions as we work together.
Step 3: Work Normally
With context loaded, proceed with whatever the user needs. The loaded memories inform your responses — you know the project's conventions, past decisions, and gotchas without the user having to re-explain them.
Tips
- Don't dump memories verbatim — absorb them and use them naturally
- Don't re-explain what the user already knows — just work with the context
- If a memory seems stale — verify against current code before acting on it
- If the user asks "what do you know" — summarize the loaded memories concisely
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.

