Skip to main content
All tools are tenant-scoped and accept ISO-8601 UTC timestamps for since / until (e.g. 2026-06-05T00:00:00Z). Prefer rollup tools (list_agents, list_agent_sessions) before paging individual decisions.

​
Discovery & rollups

​
list_agents

List agents visible to you with rollup metrics — decision count, cost, confidence, activity dates, work-pattern facts, and (when a window is given) outcome quality such as value_yield_pct and outcome_class_distribution.
​
int
default:"1"
Page number (1-based).
​
int
default:"50"
Results per page (max 500).
​
string
ISO-8601 UTC start (inclusive).
​
string
ISO-8601 UTC end (inclusive).
​
string
Comma-separated agent slugs to restrict results (max 20).
​
bool
default:"false"
Attach per-agent insight badges.

​
list_agent_sessions

List sessions for one agent, with AI session summaries, feature linkage, intent clusters, and quality metrics.
​
string
required
Agent identifier from list_agents.
​
int
default:"1"
​
int
default:"50"
Max 500.
​
string
​
string
​
string
recent (default), rework, or contradiction.
​
float
Only sessions with max contradiction ratio ≥ this (0–1).

​
list_session_decisions

List decisions in a session (summary fields only, no JSONB blobs).
​
string
required
​
int
default:"1"
​
int
default:"50"
​
string
​
string
​
string
Filter by outcome class.
​
string
Filter by work type.
​
float
​
float
​
float
​
float
​
string

​
list_decisions

Query decisions across agents and sessions with the same rich filter set.
​
int
default:"1"
​
int
default:"50"
​
string
​
string
​
string
​
string
​
string
​
string
​
float
​
float
​
float
​
float
​
string

​
get_decision

Get one decision with optional JSONB expansion. Call for a specific decision — not in a loop over a session.
​
string
required
​
string
required
UUID of the ledger row.
​
string
Comma-separated blobs: decision, metrics, reasoning_trace, tool_activity, … A good default for investigations is decision,metrics,reasoning_trace.
​
string
​
string

​
query_graph

Traverse the Enfors knowledge graph. Use only for explicit graph traversal — it can be slow (10s+). Returns [] when the graph isn’t configured.
​
string
required
One of: agent_subjects (needs agent_id), subject_decisions (needs subject_label), decision_evidence (needs ledger_id), agent_network (needs agent_id), graph_stats.
​
string
​
string
​
string
​
int
default:"20"
Max 100.
​
string
Accepted for consistency; not applied to traversals.
​
string

​
search_similar_decisions

Find decisions semantically similar to a natural-language query (vector search).
​
string
required
Natural-language description.
​
string
Restrict to one agent.
​
int
default:"20"
Max 100.
​
string
​
string

​
Delivery & ROI

​
list_features

List Features with cost / effort / rework aggregates.
​
int
default:"1"
​
string
Filter by Feature status.
​
string
​
bool
Only Features needing curation.
​
string
​
string

​
get_feature

Get one Feature’s aggregates and member sessions.
​
string
required
​
string
​
string

​
list_feature_merge_suggestions

List proposed Feature merge pairs with similarity scores.
​
int
default:"1"
​
int
default:"50"
​
string
default:"pending"

​
Insights & Pulse

​
list_insights

Return the latest insight runs with optional scope / subject filters.
​
string
entity, agent, or session.
​
string
​
string
​
bool
default:"false"
​
string
default:"summary"
​
int
default:"1"
​
int
default:"20"

​
get_pulse_report

Fetch stored Pulse report(s) for your workspace.
​
string
default:"weekly"
daily, weekly, or monthly.
​
bool
default:"true"
​
int
default:"3"
​
string
default:"summary"

​
Presentation

​
render_ui

Render Enfors data as a branded, interactive view. See render_ui.
​
string
required
HTML for the document body (the branded shell wraps it).
​
string
Optional short title for the preview tab.