Negotiation
AdvisoryAudited by Static analysis on Apr 30, 2026.
Overview
No suspicious patterns detected.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If you grant high autonomy, the agent could make counteroffers or accept a deal that affects money, work terms, or marketplace transactions.
The skill can direct the agent to communicate and, with explicit higher autonomy, accept deals. These are high-impact actions, but they are disclosed and bounded by required setup, limits, and default approval controls.
"Level 2 - Responder | Send routine replies" ... "Level 4 - Closer | Accept deals within limits autonomously" and "Default is Level 1. Never assume higher autonomy."
Use Level 1 or 2 by default, set hard limits and approval thresholds in writing, and grant Level 4 only for low-stakes categories where autonomous acceptance is truly intended.
Stored limits, salary details, offers, or negotiation patterns could reveal your leverage or be reused in later tasks if you approve retention without care.
The skill asks the agent to retain negotiation history and profile details such as limits, autonomy grants, patterns, and outcomes, which can be sensitive and can shape future negotiations.
"Log everything — Every offer, counteroffer, timestamp, who said what" and "Negotiation Profile ... Build over time. One line per insight. Confirm before storing."
Confirm exactly what should be saved, avoid storing unnecessary salary or financial leverage, and periodically review or delete outdated profile entries.