id phase-5: search by author/kind and multi-term AND are designed but not built or tracked

Finding

The design specifies search by author and by kind, and multi-term AND over tag claims, but neither is built and no issue tracks the gap. The tag-search CLI verb is wired (id claim search NAME [VALUE]), so this is purely the library-side scope expansion.

Evidence

Phase 5 record doc/2026-10-12T14-33-36Z_phase_5_tags_documents/...md, P5.5 Open: "Search by author and by kind (design §9.4) is not built. The index has no author key." and "Multi-term AND (task 06, item 3) is not built."
Plan doc/2026-10-11T11-08-42Z_design_remaining_work/plan-phases.md §10, P5.5: "Search over the tag index" — done (src/search.rs, tests/search.rs), but the design §9.4 / task 06 item 3 scope is broader than the delivered single-name search.
pkgs/id/src/search.rs: search takes one name (ValueMatch::{Any, Bare, Exact}); no author key in src/tag_index.rs (named only).

Why it matters

Plausible improvement over the delivered P5.5, already designed, so it should either be built or explicitly descoped with a note. Left untracked it will be silently dropped.

Suggested scope

Add an author key to the tag index and author/kind search terms, plus multi-term AND; or record a decision that single-name search is the intended scope and close this.

Acceptance criteria

Library search supports searching by author and by kind, in addition to the existing name/value search, and the tag index carries the author key needed for it.
Multi-term search combines terms with AND semantics.
Tests cover author search, kind search, and multi-term AND in the search / tag-index test suites; existing single-name search behavior is unchanged and its tests still pass.
OR, as an alternative closing path: a descope decision is recorded on this issue (single-name search is the intended scope) and the Phase 5 design record (P5.5) is updated to match.

CODE-119

Intake review

Related: Phase 5 record P5.5 and design §9.4 / task 06 item 3 (cited in the body); the delivered single-name tag search is pkgs/id/src/search.rs + tests/search.rs. No sibling issue tracks author/kind search or multi-term AND — this is the tracking issue for that gap.

Still live (verified 2026-10-26): no merged PR references or resolves this issue, and the body evidence stands — search still takes one name and the tag index has no author key. Not solved; do not close.

Notes / risks: This is designed-but-unbuilt scope, not a bug: the plausible outcomes are build it or explicitly descope it, and the appended acceptance criteria carry both closing paths so the item cannot sit undecided. Index changes (author key) touch the tag index format — whoever builds it should check index compatibility / rebuild behavior.

Suggested approach: Decide build-vs-descope first. If building: add the author key to the tag index, extend search terms (author, kind, multi-term AND), and test at the library level before wiring any CLI surface.

🤖 Agent note · intake enrichment

Agent: fleet monitor (repo-fleet-monitor)
Session / ID: repo-fleet-monitor run, 2026-10-26 00:19 CDT (05:19 UTC)
Found via: monitor reconciliation
Task: Intake enrichment (§5 pass): acceptance criteria appended, labels normalized, still-live verified.
Expectation: effort/l if built — tag-index + search scope expansion; the descope path is a decision only.

🤖 Agent note · picked up

Agent: queue worker (repo-worker-code), Lane A shift
Session / ID: queue task 2237 · repo-queue-worker-shift 2026-10-28 14:00 CDT
Found via: queue claim
Task: Triage issue id phase-5: search by author/kind and multi-term AND are designed but not built or tracked #800 — verify the search gap against current main and route it.
Expectation: s — done = gap confirmed at HEAD with code citations, scoped follow-up packet enqueued, verdict posted

🤖 Agent note · done

Agent: queue worker (repo-worker-code), Lane A shift
Session / ID: queue task 2237 · repo-queue-worker-shift 2026-10-28 14:00 CDT
Found via: queue claim
Task: Triage verdict for id phase-5: search by author/kind and multi-term AND are designed but not built or tracked #800 (Phase 5 search by author/kind + multi-term AND)
Expectation: s — verdict below; implementation routed as queue task 2658

Triage verdict — gap confirmed at HEAD; routed to a scoped implementation packet

Verified fresh 2026-10-28, at current HEAD:

pkgs/id/src/search.rs:50 — pub fn search(claims: &[Claim], name: &str, value: ValueMatch<'_>) -> Vec<Hit> takes exactly one name/value term; search_store (line 84) has the same single-term shape. No author term, no kind term, no multi-term composition.
pkgs/id/src/tag_index.rs:36-38 — TagIndex carries only the alpha and omega BTreeMaps; its accessors (subject, subject_name, subject_name_value, named, value, value_name, value_name_subject) are subject/name/value-shaped. There is no author key and no author accessor, even though each Entry records the claim's author. The issue's evidence is accurate in every particular.

Route: this is a real, designed, unbuilt gap — not a wontfix and not a descope candidate on the evidence (the design §9.4 scope was deliberate, and the gap is purely library-side since the id claim search CLI verb is already wired). Enqueued a scoped follow-up packet: queue task 2658 (small_fix, dedupe repo-health:v1:fix:developing-today/code:issue:800) — add the author key to the tag index, author/kind terms plus multi-term AND in search/search_store, tests for all three, existing single-name behavior unchanged. This issue stays open and is labelled queue/queued for task 2658; it closes when that implementation lands.

🤖 Agent note · parked

Agent: queue worker (repo-worker-code), Lane C shift
Session / ID: queue task 2658
Found via: queue claim (resume from awaiting_ci)
Task: parked 20m on CI — PR id(search): author index, author/kind terms, multi-term AND — fix #800 #1806 run 40804099293 still has 5 checks queued/in-progress after the 5-minute poll; clippy-lint FAILURE is infra (job log: Magic Nix Cache throttled, GitHub Actions Cache rate-limit 418 narinfo downloads, FlakeHub unauthenticated), not a lint finding; every other completed check is SUCCESS including test-unit and test-shard-0
Expectation: resume = recheck matrix; merge PR id(search): author index, author/kind terms, multi-term AND — fix #800 #1806 when green and verified, then complete task fixed_merged
Links: PR id(search): author index, author/kind terms, multi-term AND — fix #800 #1806

🤖 Agent note · parked

Agent: queue worker (repo-worker-code), lane A shift
Session / ID: queue task 2658 (resume)
Found via: queue claim (resume of parked task)
Task: Resume check on PR id(search): author index, author/kind terms, multi-term AND — fix #800 #1806 for id phase-5: search by author/kind and multi-term AND are designed but not built or tracked #800. Fresh CI: every check passes except nix flake check (clippy-lint), which failed on FlakeHub authentication infra ("FlakeHub registration required. Unable to authenticate to FlakeHub") — the same infra failure seen on PR id(world): wire manifest apply — module install + schedules over the wire (#807) #1820 this shift, now filed as ci: jobs fail with 'FlakeHub registration required' authentication error (clippy-lint observed on #1806, #1820) #2120. Re-run of failed jobs is refused while the workflow run still has a queued aggregate job.
Expectation: Parked again on CI. Resume = rerun the failed clippy job once the run completes (or once ci: jobs fail with 'FlakeHub registration required' authentication error (clippy-lint observed on #1806, #1820) #2120 is fixed), merge when green, then complete the task as fixed_merged.
Links: PR id(search): author index, author/kind terms, multi-term AND — fix #800 #1806 · branch queue/2658-800-search-author-kind · infra: ci: jobs fail with 'FlakeHub registration required' authentication error (clippy-lint observed on #1806, #1820) #2120

agent-note:v1 · agent: repo-queue-worker-shift-b (lane B) · task 2658 · kind: small_fix · outcome: picked up (resume) — PR #1806 already implements this fix (branch queue/2658-800-search-author-kind); resuming from the awaiting_ci checkpoint: verifying CI state fresh and closing out, no parallel work

agent-note:v1 · agent: repo-queue-worker-shift-b (lane B) · task 2658 · kind: small_fix · outcome: done — PR #1806 merged (merge sha 5a843a1). CI: all test jobs (test-unit, test-shard-0..3), cargo-check, doc and fmt checks SUCCESS; the only failure was clippy-lint on the FlakeHub registration/auth infra error tracked at #2120 (unrelated to this diff — search.rs, tag_index.rs, tests/search.rs only), and a rerun was refused while the aggregate run sat queued; merged under the standing unrelated-infra-failure rule with the change itself verified by the passing test/check jobs.

Navigation Menu

Repository navigation

id phase-5: search by author/kind and multi-term AND are designed but not built or tracked #800

Finding

Evidence

Why it matters

Suggested scope

Acceptance criteria

Intake review

Triage verdict — gap confirmed at HEAD; routed to a scoped implementation packet

153 remaining items

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

id phase-5: search by author/kind and multi-term AND are designed but not built or tracked #800

Description

Finding

Evidence

Why it matters

Suggested scope

Acceptance criteria

Activity

developing-today commented on Oct 24, 2026

dezren39 commented on Oct 26, 2026

Intake review

dezren39 commented on Oct 28, 2026

dezren39 commented on Oct 28, 2026

Triage verdict — gap confirmed at HEAD; routed to a scoped implementation packet

153 remaining items

dezren39 commented on Oct 29, 2026

dezren39 commented on Oct 29, 2026

dezren39 commented on Oct 29, 2026

dezren39 commented on Oct 29, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions