← Back to Skills Marketplace
rudi193-cmd

Willow Context Sentinel

by Sean Campbell · GitHub ↗ · v1.0.0 · MIT-0
linux ✓ Security Clean
69
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install willow-context-sentinel
Description
Use when you need to check whether the current session is approaching context limits and decide whether to compact, hand off, or continue. Implements a casca...
Usage Guidance
This skill appears to be what it says: a read-only local sentinel that checks two ~/.willow JSON files and prints a simple status. Before installing, consider: 1) Confirm python3 is available on your agent host (the script uses python3 but the metadata only lists bash). 2) The script intentionally 'fails open' (missing/unreadable files default to STATUS_OK) — if you prefer safer behavior, modify it to fail closed or surface an explicit error instead of allowing large operations to continue. 3) Ensure the ~/.willow files are stored with appropriate file permissions (they contain local session state). 4) The sentinel doesn't invoke remote services or exfiltrate data, but it does advise you to call other skills (/strategic-compact, /handoff); review those skills' permissions before following escalation steps.
Capability Analysis
Type: OpenClaw Skill Name: willow-context-sentinel Version: 1.0.0 The willow-context-sentinel skill is a diagnostic utility designed to help the AI agent manage its context window by monitoring local state files (~/.willow/anchor_state.json). The shell script (check_context.sh) performs read-only operations to check prompt counts and service status, using standard Python JSON parsing without any signs of data exfiltration, malicious execution, or harmful prompt injection.
Capability Assessment
Purpose & Capability
The skill is a local context sentinel that reads Willow state files under the user's home directory to decide CONTEXT_OK / COMPACT_NOW / HANDOFF_NOW / POSTGRES_DOWN. The required binary (bash) matches the runtime use. The only small mismatch is that the script invokes python3 but python3 is not declared in the metadata; otherwise the resources requested align with the described purpose.
Instruction Scope
Instructions and the script are narrowly scoped to reading ~/.willow/anchor_state.json and ~/.willow/session_anchor.json and printing a single-line status. The SKILL.md and script consistently describe read-only behavior. Note: the skill intentionally 'fails open' (missing or unreadable files default to STATUS_OK) which is an operational choice that can hide context problems; also SKILL.md examples reference calling other skills (/strategic-compact, /handoff) but the sentinel itself only outputs a status and does not invoke those skills.
Install Mechanism
Instruction-only skill with a small shipped shell script; no install spec, no downloads, and nothing written to disk beyond the included code file. Low install risk.
Credentials
The skill requests no environment variables or credentials. It reads files in $HOME/.willow, which is appropriate for a Willow-specific sentinel. One minor proportionality issue: the script uses python3 for JSON parsing but python3 is not declared as a required binary in the skill metadata.
Persistence & Privilege
always is false and the skill does not request persistent or elevated privileges, nor does it modify other skills or system configuration. Autonomous model invocation is allowed by default but that is normal and the skill's behavior is read-only.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install willow-context-sentinel
  3. After installation, invoke the skill by name or use /willow-context-sentinel
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of context sentinel cascading protocol for Willow/OpenClaw on Linux: - Monitors `prompt_count` to detect session context usage. - Outputs one of four statuses: `STATUS_OK`, `COMPACT_NOW`, `HANDOFF_NOW`, or `POSTGRES_DOWN`. - Directs automatic invocation of `strategic-compact` or `handoff` workflows when limits are reached. - Checks Postgres availability via `session_anchor.json`. - Safe, read-only operation; integrates easily with session heartbeat processes.
Metadata
Slug willow-context-sentinel
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Willow Context Sentinel?

Use when you need to check whether the current session is approaching context limits and decide whether to compact, hand off, or continue. Implements a casca... It is an AI Agent Skill for Claude Code / OpenClaw, with 69 downloads so far.

How do I install Willow Context Sentinel?

Run "/install willow-context-sentinel" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Willow Context Sentinel free?

Yes, Willow Context Sentinel is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Willow Context Sentinel support?

Willow Context Sentinel is cross-platform and runs anywhere OpenClaw / Claude Code is available (linux).

Who created Willow Context Sentinel?

It is built and maintained by Sean Campbell (@rudi193-cmd); the current version is v1.0.0.

💬 Comments