Workspace

Workspace journeys

These are the three current journeys that close the loop between governed knowledge, controlled tool use, and measurable outcomes.

Team knowledge to skilled chat

Downloads

Concepts

Journey team skill chat
Workspace current diagrams
Workspace
Journeys

Keywords

Team member
Team file store
Atlas and Lineage
Skills directory
Lens chat
upload deck or template
append provenance
verify stores:write + owning team
normalize, chunk, embed, preserve ACL
extract entities + write revision trail
install skill for team + bind store
verify member can read bound store
ask with installed skill
intersect user ACL + skill-bound stores
retrieve governed evidence
answer with [S] citations + skill version
answer/audit
Intake

Source and generation provenance

Status: current

Generated at: 2026-08-12T23:42:44.965Z

Source hash: dc38c70d6987550ce9fd46dee1f51dfdc69ade2b5e0a9addd59b4ba66860c4e7

Metadata payload hash: 881a8bf942deab6a709acc5d705f774263ec49e06f7437ef8260f48f873d1175

Canonical appearance

src/components/docs/workspace-current-diagrams.ts:35 route /docs/workspace/journeys

All appearances

canonical — src/components/docs/workspace-current-diagrams.ts:35 route /docs/workspace/journeys

No mirrored appearances.

Generation versions

App: eli-ai 0.1.0

Mermaid: 11.16.0 · Mermaid CLI: 11.16.0

Node: v26.3.1 · Yarn: 4.17.1

Renderer config hash: 68c10966fe84406ee626034d58bfabd555df9f65f691204b7c46db24038da101

Renderer theme hash: c80287a78d80ad63d27bd5ca348b2ef9a7e2f44da289e436be6484ea28a1b033

Adapter versions: diagramGenerator=2, drawioFlowchart=1, drawioGantt=1, drawioSequence=1, drawioState=1

Full sidecar JSON: team-knowledge-to-skilled-chat-dc38c70d.json

Goal: A team member uploads source material, binds an installed skill, and gets a cited answer from shared knowledge.
Current modules: Intake, Atlas, Lineage, Warrant, and Lens.
Surfaces: Team file stores, installed skills directory, and workspace chat.
Runtime sequence: Upload and ACL preservation run first, entity extraction and revision trail follow, then retrieval intersects member ACL with skill-bound stores.
Security boundary: Stores enforce owning-team and explicit grants, and skills only bind to stores the caller already reads.
Citations and audit: Lens returns source citations and appends answer provenance to lineage.
Checkpoint: Crucible acceptance-gate checks and human review queues validate skilled-chat quality before wider rollout.

Canonical deep links: Documents · Import and onboarding · Retrieval pipeline.

Governed answer with live data

Downloads

Concepts

Workspace current diagrams
Journey grounded live data
Workspace
Journeys

Keywords

freeze result snapshot + hash + data_call
Workspace user
Live data source
Lineage and Crucible
ask a question
apply live role + content ACL before ranking
retrieve certified documents and concepts
resolve entity binding + typed params
enforce entitlement + read-only gate
execute capped named query
answer, clarify, conflict, or abstain
claims with [S] and optional [D] citations
audit policy verdict + evaluation hooks
apply configured citation mode (off/shadow/enforce)
Lens
Warrant
Conduit
rows

Source and generation provenance

Status: current

Generated at: 2026-08-12T23:39:01.248Z

Source hash: eb2c637f5845cd9e283dd773c4535a01dd8d9afec528e0e2ddcba30789351d92

Metadata payload hash: 2c6a3077b4649340f72009775c3c8363614657b4615815ebf60abe65f7488a78

Canonical appearance

src/components/docs/workspace-current-diagrams.ts:54 route /docs/workspace/journeys

All appearances

canonical — src/components/docs/workspace-current-diagrams.ts:54 route /docs/workspace/journeys

No mirrored appearances.

Generation versions

App: eli-ai 0.1.0

Mermaid: 11.16.0 · Mermaid CLI: 11.16.0

Node: v26.3.1 · Yarn: 4.17.1

Renderer config hash: 68c10966fe84406ee626034d58bfabd555df9f65f691204b7c46db24038da101

Renderer theme hash: c80287a78d80ad63d27bd5ca348b2ef9a7e2f44da289e436be6484ea28a1b033

Adapter versions: diagramGenerator=2, drawioFlowchart=1, drawioGantt=1, drawioSequence=1, drawioState=1

Full sidecar JSON: governed-answer-with-live-data-eb2c637f.json

Goal: A user asks one question and receives an answer that is grounded in certified documents plus governed live data when needed.
Current modules: Lens, Warrant, Conduit, Lineage, and Crucible.
Surfaces: Workspace chat, governed live-query runner, and connected data sources.
Runtime sequence: Retrieval applies live role and content ACL before ranking, optional query execution is entitlement-gated, then policy returns answer/clarify/conflict/abstain.
Security boundary: Named query execution remains read-only, parameterized, capped, and entitlement-checked.
Citations and audit: Claims include Source [S] and optional Data [D] citations, and the configured citation mode (off, shadow, or enforce) records policy outcomes plus snapshot hash and data_call provenance.
Checkpoint: Crucible evaluation hooks and policy acceptance criteria close the feedback loop before rollout.

Canonical deep links: Answer policy · Semantic and data layer · Evaluation framework.

Capability-locked self-service agent

Downloads

Concepts

Workspace current diagrams
Journey capability locked agent
Workspace
Journeys

Keywords

Chat or MCP harness
Agent builder
Immutable version
HITL policy
run or via
invoke tool
Runner and tool registry
publish tools + capability lock + budgets
reject every tool outside the lock
exactly the admitted tool surface
score risk and resolve rule
suspend with draft + rationale
approve, reject, or edit
outcome, review verdict, citations, cost
evaluate golden set + acceptance gates + drift
log + risk-score + sample for retrospective review
tools/list
/api/mcp/agents/slug
require/autopilot
Crucible
intersect lock + key scope + live role + ACL

Source and generation provenance

Status: current

Generated at: 2026-08-12T23:37:12.800Z

Source hash: fd36754dce5d65bf61647045d41cde39a53758f65145dd50e010f06a3dba5bf5

Metadata payload hash: fe578a28dd8aea531cacacc31bdb384c9afa53de7a635a4e04070e5ea6a0682b

Canonical appearance

src/components/docs/workspace-current-diagrams.ts:76 route /docs/workspace/journeys

All appearances

canonical — src/components/docs/workspace-current-diagrams.ts:76 route /docs/workspace/journeys

No mirrored appearances.

Generation versions

App: eli-ai 0.1.0

Mermaid: 11.16.0 · Mermaid CLI: 11.16.0

Node: v26.3.1 · Yarn: 4.17.1

Renderer config hash: 68c10966fe84406ee626034d58bfabd555df9f65f691204b7c46db24038da101

Renderer theme hash: c80287a78d80ad63d27bd5ca348b2ef9a7e2f44da289e436be6484ea28a1b033

Adapter versions: diagramGenerator=2, drawioFlowchart=1, drawioGantt=1, drawioSequence=1, drawioState=1

Full sidecar JSON: capability-locked-self-service-agent-fd36754d.json

Goal: A builder publishes a reusable agent that teams can run safely through chat or an agent-scoped MCP mount.
Current modules: Ports, Warrant, Lens, Conduit, and Crucible.
Surfaces: Agent builder, immutable version views, chat harness, and agent-scoped MCP mount.
Runtime sequence: Publish sets tool allowlists and budgets, harness asks for admitted tools, invocation is risk-scored, then human review or autopilot branch executes.
Security boundary: The mount at /api/mcp/agents/{slug} exposes only the lock projection after role, key scope, and ACL intersection.
Citations and audit: Outcome, verdict, citation trail, and cost are captured for later inspection.
Checkpoint: Human approve/reject/edit controls and acceptance gates in Crucible determine whether behavior is promoted.

Canonical deep links: Creating agents · Approvals and feedback · MCP server.