Skip to content

I0501 — Group comparable work from what the records already carry

Derive the grouping that makes two pieces of work comparable from facts the records already hold, so that a recommendation compares like with like without anybody inventing a taxonomy of task types.

BLOCKED wave 0 · p3 · deep-work profile · runs alone

Part of routing-recommendations — Routing recommendations derived from measured cost.

Blocked. This issue cannot start until milestone:routing-recommendations is done. The status is derived from that, not declared.

Objective

Derive the grouping that makes two pieces of work comparable from facts the records already hold, so that a recommendation compares like with like without anybody inventing a taxonomy of task types.

Why

A recommendation is only meaningful within a class of comparable work, and the obvious way to get one is to invent categories. Invented categories are a new vocabulary nobody validates, which this repository refuses on principle and would regret in practice.

Current state

Records carry a profile, a declared scope, a verification result and an outcome. Nothing groups them, and nothing says what makes two tasks comparable.

Desired state

A derivation produces groups from recorded facts alone, states what it grouped on, and refuses to group where the facts are too thin to support it.

Scope

  • lib
  • docs

Out of scope

  • A hand-written taxonomy of task types
  • Any new noun; the grouping is a derivation, not a record

Dependencies

None. This issue is a root of the graph.

What waits on this

Acceptance criteria

  • Grouping uses only fields already recorded, and names them in its output
  • No new record type or vocabulary is introduced
  • Work that cannot be grouped is reported as ungrouped rather than forced into a group

Validation

  • bash test/run.sh

Evidence required

  • suite

Evidence

None recorded. Every token above needs a command or an artifact behind it before this issue can be completed; narrative is refused.

Risk

A grouping that is too coarse makes every recommendation the same one; too fine and no group ever reaches a usable denominator. Whichever is chosen has to be visible in the output.

Timeline

started
—
verified
—
completed
—

Those three fields, the evidence above and the state of the dependencies are all the status is made of. There is no status field to disagree with them.

Canonical record: .ai/repo/project/issues/I0501.yaml. Read it back with majordomus plan show I0501.