High-risk tools in Verimem
14 of the 237 tools in Verimem are classified as high risk. This page profiles those tools specifically, with recommended policy actions and the attack patterns that target them.
Every operation listed below is an action PolicyLayer recommends controlling at the transport layer. Open any tool to see the full profile, risk score, and YAML policy snippet.
Tools at high risk
-
hippo_compose_planExecuteFORGIA #294 — Round 10: auto-plan a multi-skill chain
-
hippo_consolidateExecuteTrigger a sleep consolidation cycle. The agent replays recent
-
hippo_curate_pipelineExecuteFORGIA #239 — full curation pipeline orchestrator.
-
hippo_decay_runExecuteCycle #110.C (2026-05-16). Run a confidence-decay pass
-
hippo_plan_stripsExecuteFORGIA #209 — Pezzo A STRIPS planner. Symbolic chaining
-
hippo_provider_switchExecuteFORGIA #201. Switch the active LLM provider at runtime
-
hippo_reasonExecuteFORGIA #212 — composite reasoning. Runs all four lenses
-
hippo_rollout_actionsExecuteFORGIA #299 — Round 15: counterfactual multi-action
-
hippo_run_taskExecuteRun a task through the HippoAgent wake loop. The agent will
-
hippo_skill_compile_macroExecuteFORGIA #253 — compile a SCHEMA-stage skill into a
-
hippo_skill_derive_predicatesExecuteFORGIA #213 — auto-derive STRIPS preconditions/post
-
hippo_skills_derive_predicates_batchExecuteFORGIA #215 — batch predicate derivation. Runs the
-
hippo_world_simulateExecuteFORGIA #289 — Round 6: world model. Predict outcome of
-
sandbox_execExecuteTask #48 — run a shell command THROUGH the deny-by-default
Attacks that target this class
High-risk tools in any server share these documented attack patterns. Each links to the full case and the defensive policy.