Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
v1.1.0 introduces cryptographic chain-of-custody verification to improve update integrity tracking. - Added chain-of-custody verification: update sequences are checked for cryptographic signing and hash chaining, detecting breaks in verifiable update history. - SKILL.md updated with new detection dimension, examples, and limitations. - Capability metadata now includes chain-of-custody-verification. - No changes to input or output format for standard trap assessments.
What This Skill Does
Detects the install-then-update attack pattern, where a skill passes initial security review at publication but later introduces malicious behavior through an automatic update that bypasses re-audit. Examines five dimensions: update policy transparency, behavioral delta on update, permission scope expansion, update timing anomalies, and rollback feasibility. v1.1 adds cryptographic chain-of-custody verification using signed, hash-chained update sequences.
It combines behavioral delta analysis, permission scope tracking, timing anomaly detection, and cryptographic chain-of-custody verification in a single report, targeting the specific install-then-update attack path rather than requiring separate tools for each dimension.
When to use it
- Auditing an installed skill before enabling automatic updates
- Comparing two specific versions of a skill for undeclared behavioral changes
- Scanning an agent's full installed skill list for combined post-install update risk
- Investigating a skill that silently expanded its permission scope after an update
- Flagging updates published during low-attention windows like weekends or holidays
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Assess install-then-update risk for data-sync-helper v1.0 through v1.2
- 1Retrieves update policy declarations across all three versions and flags silent removal of the operator confirmation requirement at v1.1
- 2Compares declared changelog entries against observed behavioral changes, identifying an undisclosed outbound connection added in v1.1
- 3Tracks permission scope across versions, noting file-read scope expansion in v1.1 and a new network-outbound permission added in v1.2
- 4Checks update publication timestamps and flags both v1.1 and v1.2 as published during low-attention windows
- 5Assesses rollback feasibility and issues a risk verdict with recommended remediation steps
Trap detection report with verdict TRAP-PATTERN-DETECTED, citing four of five trap indicators present, and recommended actions to disable automatic updates, audit outbound connections, and require explicit operator confirmation for future updates