Tools / Memory

propose_memory()

Key

Writes a memory item at status=proposed. The server auto-embeds it, links entities, runs a contradiction check against existing memory, and returns any conflicts it found. Promotion to approved is a human act.

propose_memory(content, type, entities, evidence_ids)→ PROPOSED + CONFLICTS

Callable over MCP as propose_memory(content, type, entities, evidence_ids). REST mirror POST /v1/memories Planned — alpha is MCP-only

Why your agent calls it

The agent's main write. After reading a source or finishing an analysis, it proposes what it now believes — and the server immediately tells it what that belief contradicts. Conflicts surface at write time, not at decision time, and a human decides which version of the truth gets promoted.

What it returns

propose_memory() · Propose a memory item — contradiction-checked on write.ILLUSTRATIVE

Call it from an agent

Over MCP, your agent invokes it on its own when the moment calls for it — just describe the situation. A Python SDK for headless runtimes is planned; this is what the call will look like.

agent · python sdkSDK · PLANNED — ALPHA IS MCP-ONLY

from cascadian import Cascadian

cx = Cascadian(api_key=os.environ["CASCADIAN_API_KEY"])

result = cx.propose_memory(
    content="Discount pulls deals forward; net win-rate lift unclear.",
    type="lesson",
    entities=["discount_tier", "win_rate"],
    evidence_ids=["mem_0871"],
)

Parameters

Param	Type	Description
contentREQUIRED	string	The belief, assumption, hypothesis, or lesson.
typeREQUIRED	enum	belief · assumption · hypothesis · lesson.
entities	string[]	Entities to link. The server also links automatically.
evidence_ids	string[]	Sources or facts supporting the item.

Example response

propose_memory()JSON · ILLUSTRATIVE

{
  "memory_id": "mem_0912",
  "status": "proposed",
  "entities_linked": ["discount_tier", "win_rate"],
  "conflicts": [
    { "with": "mem_0871", "kind": "contradicts",
      "summary": "existing belief says the win-rate lift is unclear" }
  ]
}

Related — Memory

keybootstrap()The ~200-token session card your agent reads first.keyingest_source()Store a raw source — chunked, embedded, indexed.keyassert_fact()Write a temporal fact with an honest available_at.keylink_memories()A typed relationship edge, with rationale.keysupersede_memory()Revise without erasing — new row, old row closed.keysearch_memory()Hybrid retrieval with an honest empty result.keyget_facts()What was true, as of any moment.keyget_entity()The versioned dossier for one entity.keyget_memory()The full evidence card, by id.keyget_context_pack()Everything knowable for a decision, stamped and typed.keylist_contradictions()Open conflicts waiting on a human.