Build a decision evolution graph from previously extracted stories. This analyzes how decisions relate across stories over time: 1. **Bucketing**: Groups stories by file overlap and keyword similarity (Union-Find) 2. **Edge classification**: Uses LLM to identify relationships between decisions i...
Part of the Kawa Code MCP MCP server. Enforce policies on this tool with Intercept, the open-source MCP proxy.
AI agents invoke evolve_decisions to trigger processes or run actions in Kawa Code MCP. Execute operations can have side effects beyond the immediate call -- triggering builds, sending notifications, or starting workflows. Rate limits and argument validation are essential to prevent runaway execution.
evolve_decisions can trigger processes with real-world consequences. An uncontrolled agent might start dozens of builds, send mass notifications, or kick off expensive compute jobs. Intercept enforces rate limits and validates arguments to keep execution within safe bounds.
Execute tools trigger processes. Rate-limit and validate arguments to prevent unintended side effects.
tools:
evolve_decisions:
rules:
- action: allow
rate_limit:
max: 10
window: 60
validate:
required_args: trueSee the full Kawa Code MCP policy for all 17 tools.
Agents calling execute-class tools like evolve_decisions have been implicated in these attack patterns. Read the full case and prevention policy for each:
Other tools in the Execute risk category across the catalogue. The same policy patterns (rate-limit, validate) apply to each.
evolve_decisions is one of the high-risk operations in Kawa Code MCP. For the full severity-focused view — only the high-risk tools with their recommended policies — see the breakdown for this server, or browse all high-risk tools across every MCP server.
Build a decision evolution graph from previously extracted stories. This analyzes how decisions relate across stories over time: 1. **Bucketing**: Groups stories by file overlap and keyword similarity (Union-Find) 2. **Edge classification**: Uses LLM to identify relationships between decisions in each bucket: - supersedes: Later decision replaces earlier (earlier is outdated) - reinforces: Later decision confirms earlier still holds - contradicts: Later decision reverses earlier - specializes: Later decision adds specificity (both remain valid) 3. **Annotation**: Labels each decision as stable, orphan, evolved, or abandoned 4. **Curation**: Keeps stable + orphan decisions, drops evolved + abandoned Note: `infer_history` already chains evolve + persist automatically. Use this tool only if you want to run evolution separately on a pre-existing set of stories. If `repoPath` is provided, curated stories are automatically persisted as intents with decisions after evolution completes. The pipeline runs asynchronously — returns immediately. Uses a cheaper model (haiku) by default since edge classification requires less reasoning than story analysis.. It is categorised as a Execute tool in the Kawa Code MCP MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Add a rule in your Intercept YAML policy under the tools section for evolve_decisions. You can allow, deny, rate-limit, or validate arguments. Then run Intercept as a proxy in front of the Kawa Code MCP MCP server.
evolve_decisions is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the evolve_decisions rule in your Intercept 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 Intercept policy for evolve_decisions. 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.
evolve_decisions is provided by the Kawa Code MCP MCP server (@kawacode/mcp). Intercept sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Open source. One binary. Zero dependencies.
npx -y @policylayer/intercept