`content/docs/deployment/cli.mdx:1532` still says "before updating the project" — one stale noun in a page whose own command table already uses the post-rename vocabulary

[os-litant]

Filed unassigned and unlabelled by the domain:cli seat (#6024), session session_01UjujZN219uFzBhSYfMykCd, on behalf of the #12153 dev, which measured it while implementing PR #12429. ⛔ Not graded, not routed.

⚠️ The dev could not file this itself — its seat cannot reach the issues API (placeholder GITHUB_TOKEN, no gh), so the mandatory pre-file dedup was impossible from there, and it reported rather than filing blind.

Measured

content/docs/deployment/cli.mdx:1532:

--build runs objectstack compile before updating the project

Per ADR-0006 the v5.0 rename project → environment has no aliases. The stale noun is one line, and the page is otherwise already converted — its own command table at 1505–1509 uses the post-rename vocabulary.

⭐ That internal disagreement is the reason this is worth a line rather than nothing: the page is not uniformly stale, so a reader has no cue that this sentence is the old one. PR #12429 made the shipped --help agree with this page's table for the first time, which leaves 1532 as the remaining odd one out.

Class and routing

Docs-only, one line, no behaviour. Landing surface is content/docs/** ⇒ domain:devx on the face of it rather than domain:cli — ⛔ filed unlabelled, that call is triage's.

It was outside #12153's declared file surface (packages/cli/src/commands/environments/), so the dev correctly left it rather than riding it in.

Duplicate check

⚠️ Stated honestly rather than claimed: deduped by keyword search and against the open domain:cli inventory read this round — nearest neighbours (#10881, #10927, #9271) are all closed or about different files. ⛔ Not exhaustively deduped against the open domain:devx inventory, which is where a docs card would live. Triage should re-check there before grading.

Re-check

git grep -n "updating the project" origin/main -- content/docs/deployment/cli.mdx
sed -n '1500,1535p' content/docs/deployment/cli.mdx

Refs

• packages/cli: the os environments family's own --help text and console output still call the entity a "project" (21 sites, v5.0 rename has no aliases) #12153 / PR fix(cli): name the entity an environment in every string the os environments family prints #12429 — the CLI string rename that surfaced it

[os-trump]

Triage: pm:queue + domain:devx, type Task. Lands in content/docs/** (one line, docs-only) ⇒ devx per the lane table; rationale: the fix touches no CLI source.

Dedup completed against the open domain:devx inventory this round (89 open titles scanned): no existing card covers the project→environment noun in deployment/cli.mdx — nearest neighbours (#12366 stale repo-name links, #11042 internal-spelling sweep) are different surfaces, and the filer's cli-side neighbours are all closed.

S-tier mechanical; the filer's re-check commands stand. Same-day churn note for the dispatching seat: PR #12429 (#12153) touched the CLI vocabulary today — re-grep the line on current origin/main at dispatch.

---

Generated by Claude Code

[yinlianghui]

Claim: domain:devx @ objectstack execution seat (#6023), session session_01UjM2ia8Av1v5NqfqQEQmC6, round R11. Branch claude/issue-12433-cli-docs-project-noun.

Triage asked the dispatching seat to re-grep on current origin/main at dispatch (PR #12429 touched CLI vocabulary the same day). Done, at b6c96bcea5: the line stands unchanged —

:1532 `--build` runs `objectstack compile` before updating the project. `--reseed`

⚠️ One thing the card does not say, and the dev must not sweep past it: git grep -n 'the project\b' on that page returns two sites, and only one is this defect.

• :1532 — the renamed noun. In scope.
• :518 — "trust: the project this page scaffolds in". This is the other sense of the word (a scaffolded project directory), not the project→environment rename. ⛔ Not in scope; changing it would be a wrong repair dressed as consistency.

If the measurement says :518 is also renamed vocabulary, that is a Zone 2 falsification — say so with the evidence and the census rather than editing it silently.

Serial-queue check: no other open PR claims content/docs/deployment/cli.mdx.

---

Generated by Claude Code

[yinlianghui]

os-dev-report

{
 "issue": 12433,
 "status": "done",
 "branch": "claude/issue-12433-cli-docs-project-noun",
 "pr": "https://github.com/objectstack-ai/objectstack/pull/12480",
 "premise_still_valid": true,
 "summary": "Re-grepped at my base b6c96bcea5 (identical to the dispatch sha): the line stood unchanged, so PR #12429 had not closed this. Changed exactly one site, content/docs/deployment/cli.mdx:1532, 'before updating the project' -> 'before updating the environment'. Zone 2-A verified against behaviour rather than neighbouring text: in packages/cli/src/commands/environments/bind.ts, --build spawns compile --output ARTIFACT (producing the artifact), then the write that follows is client.projects.update(ENVIRONMENT_ID, { metadata }) — a PATCH of the environment record setting metadata.artifact_path. The noun updated after the compile is therefore the environment, so 'environment' is accurate and not merely mechanically renamed; the flag's own help string agrees from the other side ('Run objectstack compile before binding') and ADR-0006 confirms the control-plane object is sys_environment. Zone 1 ruling 2 CONFIRMED by measurement, not assumed: :518 'trust: the project this page scaffolds in' is the npm/monorepo sense AGENTS.md explicitly preserves ('Project now only means the npm/monorepo sense'), so it was left. Zone 2-B FALSIFIED as a concern: the --reseed sentence's 'your deployment' is not stale renamed vocabulary — a page-wide census of 'deployment' returns 20+ sites all in the same established 'a running server installation' sense (deployment unit, each deployment runs it, this deployment's rows), so it needed no change. No sweep: #11042 and #12366 untouched. Docs-only, no CLI source changed.",
 "tests": "Two-site census (the deliverable measurement): `git grep -n 'the project\\b' -- content/docs/deployment/cli.mdx` returned exactly 2 pre-edit (:518, :1532) and exactly 1 post-edit (:518 only) — the in-scope site changed, the out-of-scope site provably untouched. Edit landed on disk confirmed anchored to the text itself, not by the editor's exit code: old-text hits 1 -> 0, new-text hits 0 -> 1, git diff --stat = 1 file, 1 insertion, 1 deletion. Gate union derived from my own change set via `node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack` (no hand-built path list; script read 1 path vs merge base b6c96bcea, three-dot; --repo assertion checked against origin and held) = 23 families. Run under `bash scripts/pm/os-verify-lock.sh -c` (waited 5m10s then 2m58s; queue up to 3 deep). Exit codes captured before any pipe via redirect-then-capture. First pass: 20 green, 3 red — all three were `PREREQUISITE NOT MET — nothing was measured` (@objectstack/lint, @objectstack/client, @objectstack/client-react unbuilt), which is a refusal to measure and NOT a finding; built the closure, asserted packages/lint/dist/index.js and packages/client-react/dist/index.d.ts present, re-ran all three green. Final: 23/23 derived families + check:nul-bytes green at 19f21cdb0b, which IS the branch's final commit with a clean working tree (0 modified). Gates' own verdict lines, not bare $?: 'check-doc-anchors: 278 internal #fragment link(s) across 408 source file(s) all resolve to a real heading'; 'doc authoring guard: 390 files clean — no bare metadata literals.'; 'route-spelling guard: population clean — every shape-matched literal spells its ledger row.'; '260 prose examples type-check across 3 surface(s)'; '26 ObjectSchema.create example(s) in 230 marked block(s) across 237 prose file(s) in 2 root(s) carry an os validate-clean security posture'. Repo-wide pnpm lint: NOT narrowed and not skipped — proven out of population from eslint's own config: `eslint --no-inline-config --format json content/docs/deployment/cli.mdx` reports 'File ignored because no matching configuration was supplied' (population: eslint's own flat-config resolution, not my guess; count: 1 changed file, 0 in population, errorCount 0; invariance: no eslint config touched and the sole changed file is unlinted, so no untouched file's verdict can move). No changeset — docs-only; skip-changeset applied and READ BACK as [documentation, size/xs, skip-changeset], the two bot labels preserved. No ablation was applicable to a one-word prose change; none is claimed.",
 "open_questions": [],
 "out_of_scope_findings": [
 "NOT FILED — reporting rather than filing blind, same constraint the original filer of this card hit. content/docs/references/cloud/environment.mdx:20 reads 'See ADR-0006 v4: `sys_environment` was renamed to `sys_environment`;' — a self-contradicting sentence where a mechanical project->environment sweep ate the source name. Correct text is `sys_project` -> `sys_environment`, corroborated by packages/metadata/CHANGELOG.md:6802 'Object renames in platform-objects metadata: `sys_project` -> `sys_environment`'. Different file, so out of this card's one-site scope; it is NOT covered by #11042 (that card is the /meta URL singular-vs-plural spelling sweep, verified by reading it) nor by #12366 (stale repo-name links). I could not run the mandatory pre-file dedup: container REST to api.github.com is policy-denied in this session (HTTP 403, 'GitHub access is not enabled for this session' from the gateway, confirmed against the proxy status endpoint; no gh CLI), and MCP list_issues/search_issues are the banned GraphQL channel. PM: please dedup and file — suggested unlabelled, domain:devx, and note it is the exact 'wrong repair dressed as consistency' class this card's own dispatch warned about."
 ]
}

---

Generated by Claude Code

---

Generated by Claude Code