docs(roadmap): add #467 — doctor auth preflight is Anthropic-only for OpenAI-selected models

ROADMAP.md

@@ -6422,3 +6422,75 @@ Original filing (2026-04-18): the session emitted `SessionStart hook (completed)
 
 448. **`sandbox --output-format json` has contradictory state flags — `enabled:true, supported:false, active:false, filesystem_active:true, allowed_mounts:[]`: claim that sandbox is "enabled" while OS doesn't support namespace isolation and `allowed_mounts:[]` is empty contradicts `filesystem_active:true filesystem_mode:"workspace-only"`** — dogfooded 2026-05-11 by Jobdori on `7244a82b` in response to Clawhip pinpoint nudge at `1503403842920779917` (using fresh-current-main runner at `/tmp/claw-dog-1430` per gajae's 14:00 protocol switch). Reproduction: `claw sandbox --output-format json` on macOS (where `unshare` is unavailable) returns `{"active":false,"active_namespace":false,"active_network":false,"allowed_mounts":[],"enabled":true,"fallback_reason":"namespace isolation unavailable (requires Linux with \`unshare\`)","filesystem_active":true,"filesystem_mode":"workspace-only","in_container":false,"kind":"sandbox","markers":[],"requested_namespace":true,"requested_network":false,"supported":false}`. **Three contradictions in the same envelope:** (a) `enabled:true` AND `supported:false`: what does "enabled" mean if the OS doesn't support sandboxing? Read literally, sandbox is *enabled but unsupported* — semantic nonsense. The likely intent is "user requested sandbox in config" but the field name `enabled` says "is ON". A better name would be `requested:true` or `config_intent:true`, with `enabled` reserved for the actually-active state. (b) `filesystem_active:true, filesystem_mode:"workspace-only"` AND `allowed_mounts:[]`: if the filesystem fence is active in workspace-only mode, the workspace directory itself MUST be an allowed mount. An empty `allowed_mounts:[]` array combined with `filesystem_active:true` means either (i) the fence is being misreported (it's not really active), (ii) the workspace is implicit and `allowed_mounts` only lists *additional* mounts, or (iii) the fence has no allowed paths and nothing is readable — all three are inconsistent with the user-facing summary. (c) `active:false` AND `filesystem_active:true`: the top-level `active` field is a single boolean summary, but it disagrees with `filesystem_active:true` (one component is active). Either `active` is "all components active" (then it should be `false` when any component is off) or "any component active" (then it should be `true` when filesystem is). The current value is `false` despite filesystem being active. **Sibling: no `claw sandbox --help`**: `claw sandbox status` and `claw sandbox --help` go to LLM-prompt fallback or hang (gajae confirmed at 13:00 that `sandbox status` returns typed `cli_parse` but `sandbox --help` is bounded — schema is non-uniform across help paths). **Required fix shape:** (a) rename `enabled` to `requested` or `config_intent` to disambiguate from "currently active"; (b) make `allowed_mounts` explicitly include the workspace when filesystem_mode is "workspace-only" (`allowed_mounts:[{path:"<cwd>",writable:true,reason:"workspace_root"}]`); (c) document the `active` aggregate semantics: pick either "all" or "any" composition rule and document the choice; (d) add `active_components:["filesystem"]` array as a richer alternative to the single boolean — surfaces exactly which sandbox subsystems are live; (e) regression test: when `filesystem_mode == "workspace-only"`, `allowed_mounts` MUST contain the cwd and `active` must agree with the documented composition rule. **Why this matters:** sandbox is the trust surface — automation that checks `sandbox.active == true` before running a risky LLM prompt sees `false` (no namespace, no network) and assumes no isolation, but `filesystem_active:true` means there IS partial isolation. The mixed signal forces consumers to OR all `*_active` fields together. Cross-references #428 (default permission_mode=danger-full-access — paired with sandbox-not-active means zero isolation), #444 (no broad-cwd guard — sandbox is the only safety net and its status is unclear). Source: Jobdori live dogfood, `7244a82b`, 2026-05-11.
 
+
+449. **`claw session list --output-format json` routes through `CliAction::ResumeSession` and hits the auth gate, returning `kind:"missing_credentials"` — but `session list` is a pure local filesystem read that requires no API credentials; by contrast, `claw session` (without `list`) correctly short-circuits with `kind:"unknown"` + "is a slash command" message without touching the auth gate** — dogfooded 2026-05-12 by Jobdori on `8f55870d` in response to Clawhip pinpoint nudge at `1503638404842131456`. Reproduction (no creds, isolated env): `env -i HOME=$HOME PATH=$PATH claw session list --output-format json` → `{"error":"missing Anthropic credentials...","kind":"missing_credentials"}` exit 1. `env -i HOME=$HOME PATH=$PATH claw session --output-format json` → `{"error":"`claw session` is a slash command...","kind":"unknown"}` exit 1 (no auth check). Root cause: the parser routes `session list` via `parse_resume_session_args` treating `list` as a session-path token, producing `CliAction::ResumeSession { session_path: "list", commands: [] }`. `resume_session()` then calls `LiveCli::new()` which instantiates the Anthropic client and fires the credentials guard. The `SlashCommand::Session { action: Some("list") }` special-case path in `run_resume_command()` (line 3654 comment: "`/session list` can be served from the sessions directory without a live session") is only reachable after auth passes — the no-creds guard fires before the slash-command dispatch loop. **Asymmetry:** the internal code already knows `session list` is credential-free (the comment at line 3654 says so), but the CLI entrypoint forces creds before the command ever reaches that branch. **Sibling: `session list` with no sessions returns `kind:"session_load_failed"` (from `--resume latest` fallback) rather than `{"kind":"session_list","sessions":[],"session_details":[]}` — the empty-sessions case is misrouted to the resume-failure path instead of a list-success with zero entries.** **Required fix shape:** (a) add a dedicated `CliAction::SessionList { output_format }` variant dispatched when `claw session list` is parsed — do not route through `ResumeSession`; (b) implement `run_session_list(output_format)` as a credentials-free function that calls `list_managed_sessions()` directly (same logic as the slash-command special-case at line 3659); (c) ensure empty sessions returns `{"kind":"session_list","sessions":[],"session_details":[],"active":null}` with exit 0, not a `session_load_failed` error; (d) add the same fix for sibling local-only commands that currently hit the auth gate: `session delete <id>`, `session export <id>`; (e) regression test: `claw session list --output-format json` with no credentials returns `kind:"session_list"` exit 0. **Why this matters:** session list is the canonical inventory surface for automation pipelines — `claw session list --output-format json | jq '.session_details[] | .id'` is the idiomatic way to enumerate sessions for replay, export, or resume. Requiring API credentials to read a local directory listing breaks offline use, CI environments with no API key configured, and any scripting that runs before credential setup. Cross-references #357 (session list requires creds — this is the same bug surfaced by that entry; #449 provides the root-cause path trace), #369 (session help/fork require creds), #427 (resume --help hits auth gate), #431 (skills uninstall requires creds). Source: Jobdori live dogfood, `8f55870d`, 2026-05-12.
+
+
+
+450. **`prompt` emits `kind:"missing_credentials"` JSON on STDERR (not stdout), leaving stdout at 0 bytes — automation pattern `output=$(claw prompt hello --output-format json)` captures nothing on auth-absent failure; `doctor` correctly surfaces `auth.status:"warn"` with `api_key_present:false` but exposes no `prompt_ready:false` field that automation can check before invoking `prompt`** — dogfooded 2026-05-16 by Jobdori on `a35ee9a0` in response to Clawhip pinpoint nudge at `1505208225321062521`. Exact reproduction (isolated env, no creds, fresh git repo, HEAD `a35ee9a0`): `timeout 5 env -i HOME=$ISOLATED_HOME PATH=$PATH CLAW_CONFIG_HOME=$PROBE/.claw-cfg claw prompt hello --output-format json > stdout.txt 2> stderr.txt` → stdout = **0 bytes**, stderr = 195 bytes containing `{"error":"missing Anthropic credentials…","exit_code":1,"hint":null,"kind":"missing_credentials","type":"error"}`, exit code 1. Confirms Gaebal's `1505208553793781792` pinpoint that `prompt` timeout + zero bytes was the prior state — HEAD `a35ee9a0` now correctly exits 1 with `kind:"missing_credentials"` **but the envelope is still routed to stderr** (issue #447 class, same class as prior entries #422, #435). **Contrast with `doctor`:** `claw doctor --output-format json 2>/dev/null` succeeds to stdout with `checks[auth].status:"warn"`, `api_key_present:false`, `auth_token_present:false` — but the auth check has no `prompt_ready:false` field. Automation that gates on `doctor` before invoking `prompt` must re-derive readiness from `api_key_present && auth_token_present` — there is no single canonical boolean. **Three compound problems:** (a) **stdout-empty on `--output-format json` failure**: same class as #447; `prompt`'s error envelope goes to stderr, not stdout. The canonical automation idiom `if ! result=$(claw prompt "q" --output-format json); then echo "$result" | jq .kind; fi` sees `$result=""` on failure — the jq call gets nothing. All `--output-format json` error paths must route JSON to stdout per #447 contract; (b) **`doctor` missing `prompt_ready` field**: `doctor --output-format json` already knows auth is absent (`api_key_present:false`) but surfaces no derived `prompt_ready:bool` or `prompt_blocked_reason:string` field. Automation must infer readiness from `api_key_present || auth_token_present || legacy_*_present` — a 5-field OR across legacy fields that is fragile as auth mechanisms evolve. A single `prompt_ready:false` (with `prompt_blocked_reason:"auth_missing"`) inside the `auth` check would give downstream a stable contract; (c) **`claw prompt` with no auth does no preflight and fires straight at the API**: the preflight check that `doctor` runs (auth discovery) is not reused by `prompt` to emit a fast typed error before attempting the network call. Both Gaebal's pinpoint (prompt hanging silently on older HEAD) and the current behavior (prompt hitting auth gate after a brief API attempt) stem from the same root: prompt does not short-circuit at the point where `doctor` already knows auth is absent. If `doctor` can emit `kind:"doctor"` with `auth.status:"warn"` in ~20ms without a network call, `prompt` should emit `kind:"missing_credentials"` in the same window and output it to stdout. **Required fix shape:** (a) `prompt --output-format json` must write the `kind:"missing_credentials"` JSON envelope to **stdout**, not stderr — same fix as #447 for all error envelopes; (b) add `prompt_ready:bool` and `prompt_blocked_reason:string|null` to the `auth` check in `doctor --output-format json`; derive it as `api_key_present || auth_token_present || legacy_saved_oauth_present`; (c) `prompt` must run the credential preflight check (same codepath as doctor's auth check) before attempting any API call and emit `{"kind":"missing_credentials","prompt_blocked_reason":"auth_missing"}` on **stdout** with exit 1 if the check fails; (d) `--output-format json` stdout routing fix must cover: `prompt`, `session list` (cross-ref #449), `skills uninstall` (cross-ref #431), `resume` (cross-ref #435), `acp serve` (cross-ref #443) — the full `kind:"missing_credentials"` class; (e) regression test: `claw prompt hello --output-format json` with no creds writes JSON to stdout (0 bytes stderr), exits 1, `kind:"missing_credentials"`, in under 200ms (no network attempt). **Why this matters:** `prompt` is the primary consumer entry point. Auth-absent failure routing to stderr breaks every automation wrapper that captures `$(claw prompt ... --output-format json)`. The `doctor` preflight metadata gap means auth-readiness checks require parsing 5 legacy fields instead of reading one boolean. Cross-references #447 (all JSON error envelopes on stderr), #449 (session list hits auth gate), #431 (skills uninstall hits auth gate), #357 (auth gate on local ops cluster), #422 (exit-code parity). Source: Jobdori live dogfood, `a35ee9a0`, 2026-05-16.
+
+467. **`claw doctor` auth preflight is Anthropic-only even when the selected model/provider is OpenAI-compatible: `claw --model openai/gpt-4 doctor --output-format json` with `OPENAI_API_KEY` set reports auth `warn` / `no supported auth env vars were found`, while `status` in the same invocation reports `model: "openai/gpt-4"`, `model_source: "flag"`, and `status: "ok"`. Conversely, if both `OPENAI_API_KEY` and `ANTHROPIC_API_KEY` are set with `--model openai/gpt-4`, doctor reports auth `ok` because the irrelevant Anthropic key exists, not because the selected OpenAI provider is authenticated. The preflight auth check is hardcoded to Anthropic env vars and ignores provider metadata (`api_key_env`, `base_url_env`) that runtime routing already uses** — dogfooded 2026-05-24 for the 18:00 Clawhip nudge at message `1508167732355530752`, reproduced on local `./rust/target/debug/claw` `git_sha 003b739d` (origin/main `f8e1bb72`) in a clean isolated env.
+
+    Reproduction matrix:
+
+    ```bash
+    # A. default Anthropic model + only OPENAI_API_KEY
+    env -i HOME=/tmp/iso26/home OPENAI_API_KEY=sk-openai-fake PATH=/usr/bin:/bin TERM=dumb \
+      claw --output-format json doctor | jq '.checks[] | select(.name=="auth")'
+    # => warn: no supported auth env vars were found
+
+    # B. selected OpenAI provider + only OPENAI_API_KEY
+    env -i HOME=/tmp/iso26/home OPENAI_API_KEY=sk-openai-fake PATH=/usr/bin:/bin TERM=dumb \
+      claw --output-format json --model openai/gpt-4 doctor | jq '.checks[] | select(.name=="auth")'
+    # => SAME warn: no supported auth env vars were found  ❌ wrong provider/auth preflight
+
+    # status in the same env knows the selected model is OpenAI:
+    env -i HOME=/tmp/iso26/home OPENAI_API_KEY=sk-openai-fake PATH=/usr/bin:/bin TERM=dumb \
+      claw --output-format json --model openai/gpt-4 status | jq '{status,model,model_raw,model_source}'
+    # => {"status":"ok","model":"openai/gpt-4","model_raw":"openai/gpt-4","model_source":"flag"}
+
+    # C. selected OpenAI provider + both OpenAI and Anthropic keys
+    env -i HOME=/tmp/iso26/home OPENAI_API_KEY=sk-openai-fake ANTHROPIC_API_KEY=sk-ant-fake PATH=/usr/bin:/bin TERM=dumb \
+      claw --output-format json --model openai/gpt-4 doctor | jq '.checks[] | select(.name=="auth")'
+    # => ok because ANTHROPIC_API_KEY is present, even though selected provider is OpenAI ❌
+    ```
+
+    Observed auth payloads:
+
+    ```json
+    // --model openai/gpt-4 + only OPENAI_API_KEY
+    {
+      "api_key_present": false,
+      "auth_token_present": false,
+      "details": ["Environment       api_key=absent auth_token=absent"],
+      "name": "auth",
+      "status": "warn",
+      "summary": "no supported auth env vars were found"
+    }
+
+    // --model openai/gpt-4 + OPENAI_API_KEY + ANTHROPIC_API_KEY
+    {
+      "api_key_present": true,
+      "auth_token_present": false,
+      "details": ["Environment       api_key=present auth_token=absent"],
+      "name": "auth",
+      "status": "ok",
+      "summary": "supported auth env vars are configured"
+    }
+    ```
+
+    **Root cause traced:** `rust/crates/rusty-claude-cli/src/main.rs:2071-2150` `check_auth_health()` hardcodes only Anthropic env vars:
+
+    ```rust
+    let api_key_present = env::var("ANTHROPIC_API_KEY").ok().is_some_and(|value| !value.trim().is_empty());
+    let auth_token_present = env::var("ANTHROPIC_AUTH_TOKEN").ok().is_some_and(|value| !value.trim().is_empty());
+    // summary = "supported auth env vars are configured" iff either Anthropic var exists
+    ```
+
+    But provider routing metadata already exists in `rust/crates/api/src/providers/mod.rs`: `ProviderMetadata` includes `provider`, `api_key_env`, `base_url_env`, `default_base_url`, and model-prefix detection. Runtime dispatch uses `detect_provider_kind(&model)` / metadata to route OpenAI-compatible models; doctor/auth preflight ignores it and evaluates Anthropic only. This creates a split-brain preflight: `status` sees the OpenAI model flag, runtime would use OpenAI-compatible auth, but `doctor` audits Anthropic env vars.
+
+    **Why distinct from existing items:** ROADMAP **#28** improved auth error copy when users set adjacent provider env vars or put `sk-ant-*` in the wrong Anthropic var. This pinpoint is not error copy during a failed API call; it is local `doctor` preflight using the wrong provider's auth rules. ROADMAP **#29** fixed CLI runtime dispatch so `openai/...` uses OpenAI-compatible client; this entry shows the preflight health check did not follow that routing fix. ROADMAP **#465** covers Anthropic-only mixed env vars (`ANTHROPIC_API_KEY` + `ANTHROPIC_AUTH_TOKEN`) and missing `effective_auth_source`; this entry covers cross-provider model-selected auth (`OPENAI_API_KEY` should be considered when selected provider is OpenAI). ROADMAP **#466** covers provider base URL validation; this covers provider auth env validation. ROADMAP **#248** asks runtime lifecycle events to expose provider/auth source after prompt starts; this is the local doctor/status preflight counterpart before prompt starts.
+
+    **Why this matters:** (1) **False warning:** OpenAI users following documented prefix routing (`--model openai/gpt-4` + `OPENAI_API_KEY`) get `doctor` auth warn, even though the selected provider has the expected key. That makes doctor untrustworthy for non-Anthropic lanes. (2) **False ok:** If a stale Anthropic key exists in the environment, doctor reports auth ok for an OpenAI model even if `OPENAI_API_KEY` is missing. That is worse than no check because it green-lights a lane that will fail with missing OpenAI credentials. (3) **Routing and preflight drift:** #29 fixed runtime dispatch, but doctor still reasons as if claw is Anthropic-only. Every new provider added to metadata will inherit this preflight false-negative/false-positive unless doctor keys off provider metadata. (4) **Automation cannot trust auth preflight.** A clawhip lane planner checking `doctor` before starting an OpenAI-compatible prompt cannot tell whether the lane is actually authenticated for the selected provider. (5) **The data model already exists.** Provider metadata carries the exact env var names; the diagnostic layer just needs to use it.
+
+    **Required fix shape:** (a) Thread the selected model/provider into `check_auth_health()` (or add a `check_provider_auth_health(model)` helper) and resolve `ProviderMetadata` with the same function runtime dispatch uses. (b) Emit structured fields: `selected_provider`, `required_api_key_env`, `required_auth_envs`, `selected_provider_api_key_present`, `anthropic_api_key_present`, `openai_api_key_present`, `xai_api_key_present`, `dashscope_api_key_present`, and `effective_auth_source` where applicable. (c) For OpenAI-compatible selected providers, auth status should be `ok` when the provider's own key is present, regardless of Anthropic vars; warn/fail when it is absent even if Anthropic keys exist. (d) Mirror provider-auth summary into `status --output-format json`, since status is the lightweight preflight surface. (e) Regression matrix: default Anthropic + Anthropic key; OpenAI model + OpenAI key; OpenAI model + only Anthropic key (should warn/fail); OpenAI model + both keys (should ok with selected_provider=`openai`, not because Anthropic exists); XAI/DashScope equivalents. **Acceptance check:** `env -i HOME=$TMP OPENAI_API_KEY=sk-test PATH=$PATH claw --model openai/gpt-4 doctor --output-format json | jq -e '.checks[] | select(.name=="auth") | .status == "ok" and .selected_provider == "openai" and .required_api_key_env == "OPENAI_API_KEY"'` should pass; currently it warns that no supported auth env vars were found. Source: gaebal-gajae dogfood for the 2026-05-24 18:00 Clawhip nudge. Coordination note: still avoided F/CLAW_CONFIG_HOME due to Jobdori public claim; this provider-auth-preflight mismatch is orthogonal and credential-free.