engram_startStart a new agent session with one canonical call.
Call first after setup/restart, before the agent starts making product or code decisions.
session_idExact journal id from start/rehydrate/list_sessions. Disambiguates parallel work; writing requires your active session. Rehydrate may read shared or closed history.
agent_toolCalling client name for attribution, e.g. codex, claude-code, cursor, or gemini. Omit if unknown.
projectProduct/project name for session context. Start/setup use it as the repository hint when repo is omitted.
repoRepository slug, e.g. engram or billing-api, within the authenticated workspace. This is a memory filter, not a filesystem path or workspace selector.
roleAgent responsibility used in task context, e.g. backend engineer. Start/setup default to implementation agent.
marketBusiness/domain context, e.g. fintech or education; not a geographic access setting.
taskConcrete current goal or decision, e.g. diagnose checkout retries; used to focus context and session recovery.
stackProject technologies used to focus context; omit when unknown.
contextOptional full task context; overrides the query generated from project, role, market, stack and task.
new_sessionDefault false reuses one eligible active journal. True creates a separate journal for parallel work; retries can create another session. Keep the returned session_id.
include_skillsInclude relevant procedural checklists. If omitted, the server infers this from task/context keywords; false suppresses checklists.
limitMaximum retrieved memories, 1β30. Omitted defaults: focused=8, deep=16; explicit values override these defaults.
depthDefault focused uses a small retrieval budget; deep expands it when focused results are insufficient. An explicit limit overrides the default budget.
pending_feedbackBatch feedback for earlier retrieval_ids, committed independently before this action. Requires outcome:write; same actual-use feedback is safe to retry. No extra call quota. Omit unassessed results.