Query the compensation-signatures lens: how stereotyped is your chain-shape behavior, and which chain patterns dominate vs which are exploratory? Every multi-step chain the agent runs is fingerprinted as an ordered target sequence (e.g. api:openai.com → api:stripe.com → mcp:filesystem). Across a ...
Risk signalsBulk/mass operation — affects multiple targets
Part of the ACR — Agent Composition Records server.
Free to start. No card required.
AI agents call get_compensation_signatures to retrieve information from ACR — Agent Composition Records without modifying any data. This is common in research, monitoring, and reporting workflows where the agent needs context before taking action. Because read operations don't change state, they are generally safe to allow without restrictions -- but you may still want rate limits to control API costs.
Even though get_compensation_signatures only reads data, uncontrolled read access can leak sensitive information or rack up API costs. An agent caught in a retry loop could make thousands of calls per minute. A rate limit gives you a safety net without blocking legitimate use.
Read-only tools are safe to allow by default. No rate limit needed unless you want to control costs.
{
"version": "1",
"default": "deny",
"tools": {
"get_compensation_signatures": {}
}
}See the full ACR — Agent Composition Records policy for all 30 tools.
These attack patterns abuse exactly the kind of access get_compensation_signatures gives an agent. Each links to the full case and the policy that stops it:
Other read tools across the catalogue. The same approach applies to each: allow, with a rate cap to control cost.
Query the compensation-signatures lens: how stereotyped is your chain-shape behavior, and which chain patterns dominate vs which are exploratory? Every multi-step chain the agent runs is fingerprinted as an ordered target sequence (e.g. api:openai.com → api:stripe.com → mcp:filesystem). Across a window, the lens reports: • agent_stability — a continuum score in [0, 1]. 1.0 = one pattern does everything (maximally routine). 0.0 = every chain looks different (exploratory / unstable). Computed as 1 − normalized Shannon entropy across patterns. • pattern_stability — per-pattern share of total chains. A high value means this exact sequence is the agent's routine. A low value with persistent frequency is the kind of long-tail signal that *can* be compensation (routing around something) or genuinely exploratory — you read it together with the friction report. • fleet_agent_count — how many other agents run this same pattern. A high-frequency, low-fleet pattern is idiosyncratic. A fleet-wide pattern is a substrate-level signal. This is a continuum, not a verdict. There is no "compensation detected" flag — only the distribution. Interpret a persistent low-stability tail with non-trivial frequency as *possible* ongoing compensation, and confirm by cross-referencing the friction report for the targets involved. Requires at least some multi-step chains to have been logged with chain_id + chain_position. Window is 'day' or 'week'; defaults to 'week'. Runs against chain_analysis, which is refreshed nightly by the background job.. It is categorised as a Read tool in the ACR — Agent Composition Records MCP Server, which means it retrieves data without modifying state.
Register the ACR — Agent Composition Records MCP server in PolicyLayer and add a rule for get_compensation_signatures: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches ACR — Agent Composition Records. Nothing to install.
get_compensation_signatures is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the get_compensation_signatures rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for get_compensation_signatures. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
get_compensation_signatures is provided by the ACR — Agent Composition Records MCP server (@tethral/acr-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 30 ACR — Agent Composition Records tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4,600+ MCP servers and 31,000+ tools scanned and risk-classified.