ChronovelChronovel

What is Chronovel?

Chronovel is a Decision Accountability Platform. It makes decisions first-class objects in your organization: authored, timestamped, linked to an org-chart hierarchy, and tied to customer outcomes via AI attribution.

Think of it as Git for organizational decisions. Every decision has an author, a timestamp, a context, and an outcome. Anyone in the org can trace any customer result back to the guidance that shaped it.

The two core problems Chronovel solves:

  • –Decisions are invisible — made in meetings, never written down, impossible to trace.
  • –ICs don't know what they're allowed to decide — so they interrupt managers or make silent choices.

Quickstart

Get up and running in under 10 minutes.

1. Create your account

Go to app.chronovel.com and register with your work email. The first account in an org automatically becomes the root node (CEO / team lead).

2. Set up your org chart

Navigate to Org Chart and create the first node for yourself. Then invite teammates — each person gets a node linked to their account. Position each node under its correct parent to reflect the real reporting structure.

3. Write your first guidance

Go to Guidance Feed → New guidance. Give it a title, write the body in Markdown, choose a scope and visibility, and save as draft. When you're ready, publish it — it becomes active and visible to everyone in your subtree.

4. Log a decision

When you make a call — large or small — log it as a Decision: give it a title, describe what you chose and why, and link it to any relevant context (ticket, project, initiative). If your decision deviates from an existing guidance, select the guidance and provide a written rationale. Every decision is timestamped, attributed, and append-only.

5. Request guidance when uncertain

If no guidance covers a decision you need to make, open a Guidance Request and route it to the appropriate ancestor. This creates a visible, tracked obligation for them to respond.

Setting up your org chart

Chronovel models your organization as a tree. Each node represents one person in one role. Nodes inherit all guidance from their ancestors — the closer to the root, the broader the reach.

The first node created has no parent and becomes the root. Only one root per org is allowed. Subsequent nodes must be placed under an existing node.

Key fields per node:

  • –Role — the person's title (e.g. VP Engineering)
  • –Department — used for filtering and reporting
  • –Can produce guidance — whether this person is authorized to publish guidance

To move a node, use Move from the org chart view. Chronovel automatically recalculates the ancestor path for the moved node and all its descendants, so visibility stays correct without manual intervention.

Guidance

A guidance is a named rule, directive, or constraint. It is written by someone with authority, scoped to their domain, and visible to everyone below them in the org chart. Guidance tells people what frame to operate in — it constrains future decisions without making them.

Lifecycle

  • –draft — saved but not visible to the team. Edit freely.
  • –active — published and visible to the org subtree. Body is immutable.
  • –superseded — replaced by a newer guidance. Linked to its successor.
  • –withdrawn — deactivated. No longer appears in feeds.

Scope

Each guidance has a scope that describes the kind of decision it governs:technical, process, product, people, or business. Scopes are used for filtering in the feed and for AI attribution.

Visibility

  • –team — visible only to direct reports of the author's node
  • –department — visible to the author's department subtree
  • –company — visible to the entire org

Decisions

A decision is a logged choice. Where guidance sets the frame, decisions fill it — they record what was chosen, in what context, and why. Decisions are the unit of accountability in Chronovel.

Anyone in the org can log a decision. It does not require authority — it requires honesty. The decision log is append-only: final decisions are never edited or deleted.

Anatomy of a decision

  • –Title — a short label for the choice made.
  • –Body — what was decided and why, in Markdown.
  • –Context — optional link to a ticket, project, or initiative.
  • –Deviations — any active guidances this decision consciously departs from, each with a required written rationale.
  • –Status — draft (editable) or final (immutable).

Deviations

If your decision departs from an existing active guidance, you must name the guidance and explain why. This is not punitive — it is data. A well-documented deviation is better than a silent one. Deviations feed the AI attribution model and help leadership understand where guidance needs to evolve.

Following guidance requires no special action. If you make a decision that is fully within the existing guidance frame, just log the decision — no deviation record needed.

Org hierarchy and visibility

Chronovel stores the org tree as an adjacency list with materialized paths. Each node holds an ancestorIds array — the ordered list of all nodes from root to parent. This allows all hierarchy queries to be single indexed reads with no recursion.

What you can see

You see all active guidances from your own node and every ancestor. You never see guidances from sibling branches or descendant-only nodes.

Who you can request from

Guidance requests can only be routed to ancestors — nodes above you in the hierarchy. You cannot request from peers or from other branches.

Outcomes and attribution

Outcomes are customer-facing signals ingested into Chronovel: NPS scores, CSAT, support ticket volume, churn events, revenue changes. Each outcome is tagged with a period (e.g.2026-Q1) and then attributed to the decisions and guidances that plausibly influenced it.

Attribution is AI-generated (Claude) and always includes a confidence score and a written rationale. It is a recommendation layer — humans make the final call on performance reviews and bonuses.

Guidance agent

The guidance agent is a conversational interface that surfaces relevant guidances from your org hierarchy in response to natural language questions.

How it works

  • –Your query is embedded and matched against all active guidances visible to you.
  • –Results are ranked by semantic similarity and returned with full attribution (author, role, date).
  • –The agent never answers from its training data — if no guidance exists, it says so.
  • –You can ask follow-up questions within the same session.

Example

User: How should we handle database migrations in production?

Agent: I found 2 relevant guidances:

1. "Database migrations must be backward-compatible"
   Sofia Kim, Backend Lead — 20 Jan 2026
   Relevance: directly covers migration strategy.

2. "PRs require approval from at least one backend lead"
   Sofia Kim, Backend Lead — 15 Jan 2026
   Relevance: applies to the PR containing the migration.

No guidance exists on rollback procedures. Would you like to
escalate a request to Sofia Kim?

Task interpretation

Task interpretation is a structured-output mode of the guidance agent. Paste any task, ticket, or decision description and receive a structured breakdown of your constraints.

The response has three sections:

  • –Must follow — active guidances that directly constrain this task, with author and rationale.
  • –You decide — decision points in the task that no guidance covers. You own these calls.
  • –Escalate — gaps significant enough to warrant a new guidance request to an ancestor.
On the free tier, paste tasks manually. On Pro, the native Jira integration fetches tickets automatically — no copy-paste required.

Export for AI

Available on all tiers. The Export for AI button sends your org's decision data — guidances, actions, outcomes — to Claude for one-off analysis. Each call is billed at cost via Anthropic with no markup.

Use cases:

  • –Outcome attribution for a specific period
  • –Guidance quality and clarity feedback
  • –Identifying gaps in your decision coverage
  • –Promotion and bonus recommendation drafts

Outcome attribution

Quarterly, Chronovel runs an attribution pass over your active guidances and ingested outcomes. Claude reads both and returns a ranked list of guidances that plausibly contributed to each outcome, with a confidence score (0–1) and a written explanation.

Attribution drives two downstream reports:

  • –Per-person guidance production score — how much active, outcome-linked guidance did this person author?
  • –Promotion and bonus recommendations — a ranked summary with rationale, reviewed by leadership before any action.
Attribution is a signal, not a verdict. Chronovel surfaces the data; your leadership team makes the decisions. The model is explicit about confidence and never hides uncertainty.

Jira integration (Phase 2)

The Jira integration connects Chronovel's task interpretation feature directly to your Jira board. Instead of manually pasting ticket text, the agent fetches the ticket by ID and runs interpretation automatically.

  • –OAuth2 connection to your Jira workspace
  • –Fetch any ticket by key (e.g. ENG-4821)
  • –Interpretation results posted back as a Jira comment (optional)
  • –Available on Pro and Enterprise plans

Phase 2 is on the roadmap. Free tier users can use task interpretation via manual paste in the meantime.

REST API reference

All Chronovel features are available via REST API. Authentication uses JWT Bearer tokens issued at login. Access tokens expire after 15 minutes; use the refresh endpoint to rotate.

Authentication

POST /auth/login
{ "email": "...", "password": "..." }
→ { "accessToken": "...", "user": { ... } }

POST /auth/refresh   # uses httpOnly cookie
→ { "accessToken": "..." }

POST /auth/logout

Guidance

GET    /guidance              # feed visible to current user
GET    /guidance/mine         # guidances authored by current user
GET    /guidance/:id
POST   /guidance              # create draft
PATCH  /guidance/:id          # update (draft only)
POST   /guidance/:id/publish
POST   /guidance/:id/withdraw
POST   /guidance/:id/supersede

Decisions

GET    /decisions             # decisions visible to current user
GET    /decisions/mine        # decisions authored by current user
GET    /decisions/:id
POST   /decisions             # create (status: draft)
PATCH  /decisions/:id         # update (draft only)
POST   /decisions/:id/finalize

Org chart

GET    /org/tree
GET    /org/me
GET    /org/nodes/:id
GET    /org/nodes/:id/ancestors
GET    /org/nodes/:id/subtree
POST   /org/nodes
PATCH  /org/nodes/:id
POST   /org/nodes/:id/move
DELETE /org/nodes/:id

Guidance requests

GET    /requests          # requests targeting me
GET    /requests/sent
GET    /requests/:id
POST   /requests
PATCH  /requests/:id     # action: answer | escalate | close

Webhooks

Webhook support is on the roadmap for Phase 2. Planned events:

  • –guidance.published — a new guidance became active
  • –guidance.superseded — a guidance was replaced by a successor
  • –decision.created — a decision was finalized
  • –request.created — a guidance request was opened
  • –request.answered — a guidance request received a response
  • –attribution.complete — an AI attribution run finished

Until webhooks ship, real-time updates are available via the Socket.IO connection used by the web app. Contact us if you need webhook access before the official release.