OpenClaw plugin logs when an agent calls a tool but still misses the real outcome
blakemcthe27 · reddit · 2026-07-29
OpenClaw plugin tracks when an agent succeeds at a tool call but still gets the outcome wrong
The author launched McPherson Governance v0.5.1, the first public plugin in the broader Observa platform.
The problem it targets
An activity log can show that an agent called a tool, but it does not prove that:
- the action was authorized
- the external system changed as intended
- the result was verified
- partially completed work was recovered
- unresolved work was handed back to a human
What the plugin does
The plugin runs alongside an OpenClaw agent and records:
- attempted actions
- shadow policy evaluations
- evidence status
- review-required items
- unresolved outcomes
It is intentionally non-enforcing: it cannot block, approve, deny, or rewrite actions. The author says they started in shadow mode because enforcement should not be turned on based on theoretical policy alone; first you need to observe real workflows and identify actual risk boundaries.
Early results and next step
- The first public release had clean-install issues, which were patched in v0.5.1.
- In a little over 48 hours, the repo recorded 21 clones from 18 unique cloners.
- The author is looking for a small number of operators to test it on one bounded workflow, such as a CRM update, outbound message, scheduling, refund flow, or another multi-system action.
The core question posed to readers: what would a governance layer need to prove before you’d trust it around a real side effect?
More from coding & agent
- ChatGPT vs Claude Browser Agents: Login Persistence Makes the Difference — dkundel · 2026-07-30
- The Worst Enterprise AI Retrieval Bug: Confident but Incomplete Answers — rohanpaul_ai · 2026-07-30
- Pulse brings local network logging and request inspection to Apple apps — tom_doerr · 2026-07-30
- Rust terminal diff viewer adds AI commit messages and PR review — iamsahaj_xyz · 2026-07-30
- OpenSquilla v0.5.1 splits model access from routing and claims big cost savings — hey_abusiddik · 2026-07-30
- Google ADK article argues tool-call interceptors are the missing safety layer for agents — fhinkel · 2026-07-30