F1a: repoint deleted-TASKS.md bind cites in std/compiler/lens to live concept issues (comment-only)

[briansrls]

What

Comment-only PR (1 of ~3 in the F1a lane). src/v4/TASKS.md was deleted 2026-06-01 in #4192, leaving 🟡 marks bound to a dead ledger. This PR repoints every TASKS.md cite under src/v4/std/, src/v4/compiler/, and src/v4/lens/ to its dissolve-on concept tracking issue, one issue per concept (the F1a mechanism from docs/planning/label-hygiene-census-2026-06-12.md, PR #4749; pattern follows #4750/#4751):

• v4 claim eval runner: Outcome-carrying eval replaces structural projection and Outcome→Bool claim bridges (re-anchor of deleted TASKS.md T-22 binds) #4757 claim eval runner (Outcome eval replaces structural/Bool bridges) — former T-22
• content-hash & cache-key boundary: bounded ByteOffset carrier, content_hash(Node) digests, affected-set pruning, CI selection (re-anchor of deleted TASKS.md T-21/T-22/T-24 + IRT-4 binds) #4758 content-hash & cache-key boundary — former T-21/T-22/T-24/IRT-4
• coproduct-arm reflection replaces hand-enumerated variant projections (re-anchor of deleted TASKS.md T-3/T-19 binds) #4759 coproduct-arm reflection — former T-3/T-19
• infer-authored grounding replaces hand-enumerated fixture facts (re-anchor of deleted TASKS.md T-9/T-10 binds) #4760 infer-authored grounding — former T-9/T-10
• grammar/lex substrate carriers: epsilon/sequence shims, LexRule/LexCharClass projections, typed grammar-relation rows, layout facts (re-anchor of deleted TASKS.md T-6/T-7/T-11 binds) #4761 grammar/lex substrate carriers — former T-6/T-7/T-11
• lens application + testgen canonicalization: CompileLens locals, coverage-key projections, LeafModelFixture rows, toolchain coproducts (re-anchor of deleted TASKS.md T-4/T-19/T-22-follow-up binds) #4762 lens application + testgen canonicalization — former T-4/T-23
• ModelCore Effect/PartialitySemanticsDecl population (re-anchor of merged PR #3474 / deleted TASKS.md T-33 binds) #4764 ModelCore Effect/PartialitySemanticsDecl — former T-33 (merged Wave-4-B T-33 — std/model_core.dag substrate (Ratified Q1; shared substrate factoring; LanguageModel + HostModel both extend it; primitive types, algebra inhabitance, laws, effect semantics, partiality) #3474 was a dead anchor)
• LanguageModel carrier ownership: surface_atom identity lookup + Set<Symbol> membership reification (re-anchor of deleted TASKS.md T-8 binds) #4766 LanguageModel carrier ownership / Set reification — former T-8

Each issue restates the dissolve-on conditions verbatim and lists the bound sites. Dead-link Anchor: header cites (fact_density.dag pointing at the deleted TASKS.md URL; coercion/constraints/lexing/refinement header cites) rewritten or dropped; algebra.dag keeps its live node:// bind and only loses the dead (ledger: T-2 · TASKS.md §T-2) tail.

Coverage note (honest skips)

Files touched by open PRs are deliberately NOT touched here, deferred to a follow-up pass after they merge: src/v4/std/compilers/target_model.dag, src/v4/compiler/04_infer.dag, src/v4/compiler/06_value_expression.dag (no TASKS.md cites today anyway), and src/v4/extdeps/languages/dag.dag (5 cites → #4761/#4763) + typescript.dag (2 cites → #4760/#4761), both in #4741's diff.

Remaining TASKS.md cites in extdeps/ and test/claim/ land in the next two small PRs of this lane.

Verification

grep -rn 'TASKS\.md' src/v4/std/ src/v4/compiler/ src/v4/lens/ --include='*.dag' now returns only re-anchor/former ledger annotations. Diff is comments only — no code, no identifiers, no atoms.

🤖 Generated with Claude Code

[briansrls]

Pushed a coupling fix: v4_test_bootstrap_infra_closeout_test.rs:584 asserts the verbatim bound task: src/v4/TASKS.md#t-19-… text of the verification.dag mark this PR rewrites — the assert literal is updated in the same PR (same-PR coupling rule), so the diff is no longer strictly comment-only: it includes that one test string literal. — sent from vivid-cat-408

[briansrls]

The lexing.dag:77 §T-6).1 mangle (flagged by both claude and cursor reviews) is fixed — but by deletion rather than repair: the operator ruled mid-review that the historical "(re-anchor of deleted TASKS.md …)" parentheticals should not appear in code comments at all. The latest commit strips them across the PR; marks now carry just the live bind (e.g. bound task: gunbc#4761), and the deleted-ledger genealogy lives in the tracking issues themselves (#4757–#4766). The coupled closeout-test literal moved in step; check_t19_testgen_activation passes on this tip, and the integration closeout suite matches the main baseline (same 4 pre-existing failures, no new ones). — sent from vivid-cat-408