Ctx UpgradeSAFE

Context window optimization for AI coding agents. Sandboxes tool output (98% reduction), persists session memory, and enforces routing across 17 platforms via MCP + hooks.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: ctx-upgrade
description: |
  Update context-mode from GitHub and fix hooks/settings.
  Pulls latest, builds, installs, updates npm global, configures hooks.
  Trigger: /context-mode:ctx-upgrade
user-invocable: true
---

# Context Mode Upgrade

Pull latest from GitHub and reinstall the plugin.

## Instructions

1. Call the `ctx_upgrade` MCP tool directly. It returns a shell command to execute.
2. Run the returned command using your shell execution tool (Bash, shell_execute, etc.).
3. Display results as a markdown checklist:
   ```
   ## context-mode upgrade
   - [x] Pulled latest from GitHub
   - [x] Built and installed v1.0.39
   - [x] Hooks configured
   - [x] Doctor: all checks PASS
   ```
   Use `[x]` for success, `[ ]` for failure. Show actual version numbers.
4. Tell the user to **restart their session** to pick up the new version.
5. **Fallback** (only if MCP tool call fails): Derive the **plugin root** from this skill's base directory (go up 2 levels — remove `/skills/ctx-upgrade`), then run with Bash:
   ```
   CLI="<PLUGIN_ROOT>/cli.bundle.mjs"; [ ! -f "$CLI" ] && CLI="<PLUGIN_ROOT>/build/cli.js"; node "$CLI" upgrade
   ```

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.