diff --git a/.github/scripts/feature-ideation/lint-caller.sh b/.github/scripts/feature-ideation/lint-caller.sh
new file mode 100755
index 000000000..eab96c1fe
--- /dev/null
+++ b/.github/scripts/feature-ideation/lint-caller.sh
@@ -0,0 +1,151 @@
+#!/usr/bin/env bash
+# lint-caller.sh — guard feature-ideation caller stubs against the #571
+# zero-job startup failure.
+#
+# Why this exists:
+#   A reusable-workflow call graph (`jobs.<id>.uses` + its `with:`) is validated
+#   at WORKFLOW SETUP — before, and regardless of, the calling job's `if:`. The
+#   `inputs` context is only populated for `workflow_dispatch` / `workflow_call`
+#   events, so a `with:` value that references `${{ inputs.* }}` fails the whole
+#   run (zero jobs, "Invalid workflow file") when the workflow is also triggered
+#   by an event that has no `inputs` — e.g. `discussion: created`. The job `if:`
+#   does NOT save you. This is petry-projects/.github#571.
+#
+#   The fix is to resolve dispatch inputs in an ordinary `prep` job (whose step
+#   expressions run at job time and are skipped on `discussion`) and pass them to
+#   the reusable call via `needs.prep.outputs.*` — an always-valid context that
+#   defers the `with:` evaluation to run time.
+#
+# This linter flags any job-level reusable `uses:` whose `with:` references the
+# `inputs` context. `github.event.inputs.*` and step-level action `uses:` are
+# intentionally NOT flagged.
+#
+# Usage:
+#   lint-caller.sh [<workflow.yml> ...]
+#   With no args, scans the two feature-ideation caller stubs.
+#
+# Exit codes:
+#   0  no issues
+#   1  one or more findings
+#   2  bad usage / file or parse error
+
+set -euo pipefail
+
+scan_file() {
+  local file="$1"
+
+  python3 - "$file" <<'PY'
+import re
+import sys
+
+try:
+    import yaml
+except ImportError:
+    sys.stderr.write("[lint-caller] PyYAML is required (pip install pyyaml)\n")
+    sys.exit(2)
+
+path = sys.argv[1]
+try:
+    with open(path, "r", encoding="utf-8") as f:
+        doc = yaml.safe_load(f)
+except (OSError, yaml.YAMLError) as exc:
+    sys.stderr.write(f"[lint-caller] cannot parse {path}: {exc}\n")
+    sys.exit(2)
+
+if not isinstance(doc, dict):
+    sys.exit(0)
+
+jobs = doc.get("jobs")
+if not isinstance(jobs, dict):
+    sys.exit(0)
+
+# `inputs` used as a context ROOT — i.e. `inputs.foo` or `inputs['foo']` — but
+# NOT `github.event.inputs.foo` (where `inputs` is preceded by a `.`). The
+# negative lookbehind excludes a leading `.` or word character.
+inputs_ctx = re.compile(r"(?<![.\w])inputs\s*[.\[]")
+# Non-greedy so adjacent ${{ }} expressions on one line are matched separately.
+gh_expr = re.compile(r"\$\{\{(.*?)\}\}", re.DOTALL)
+
+
+def flatten(value):
+    """Yield every scalar string reachable from a `with:` value."""
+    if isinstance(value, str):
+        yield value
+    elif isinstance(value, dict):
+        for v in value.values():
+            yield from flatten(v)
+    elif isinstance(value, list):
+        for v in value:
+            yield from flatten(v)
+
+
+findings = []
+for job_id, job in jobs.items():
+    if not isinstance(job, dict):
+        continue
+    # Only reusable-workflow calls: a job with a top-level `uses:` key. Step-level
+    # `uses:` lives under `steps:` and is never inspected here.
+    if "uses" not in job:
+        continue
+    with_block = job.get("with")
+    if not isinstance(with_block, dict):
+        continue
+    for key, value in with_block.items():
+        flagged = False
+        for scalar in flatten(value):
+            for expr in gh_expr.findall(scalar):
+                if inputs_ctx.search(expr):
+                    findings.append((job_id, key, scalar.strip()))
+                    flagged = True
+                    break
+            if flagged:
+                break
+
+if findings:
+    sys.stderr.write(
+        f"[lint-caller] {len(findings)} reusable `with:` value(s) in {path} "
+        f"reference the `inputs` context.\n"
+        f"  On a `discussion` (or other non-dispatch) trigger this fails at "
+        f"workflow setup with zero jobs (#571), regardless of the job `if:`.\n"
+        f"  Resolve inputs in a `prep` job and pass via `needs.prep.outputs.*`.\n"
+    )
+    for job_id, key, scalar in findings:
+        sys.stderr.write(f"  job '{job_id}', with.{key}: {scalar}\n")
+    sys.exit(1)
+sys.exit(0)
+PY
+  return $?
+}
+
+main() {
+  if [ "$#" -eq 0 ]; then
+    # Default: the two feature-ideation caller stubs (template + this repo's own).
+    local repo_root
+    repo_root="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")/../../.." && pwd)"
+    set -- \
+      "${repo_root}/standards/workflows/feature-ideation.yml" \
+      "${repo_root}/.github/workflows/feature-ideation.yml"
+  fi
+
+  local exit=0
+  local file_rc=0
+  for file in "$@"; do
+    if [ ! -f "$file" ]; then
+      printf '[lint-caller] not found: %s\n' "$file" >&2
+      exit=2
+      continue
+    fi
+    # Preserve exit-2 (file/parse error) over exit-1 (lint finding), matching
+    # lint-prompt.sh's precedence.
+    scan_file "$file" && file_rc=0 || file_rc=$?
+    case "$file_rc" in
+      0) ;;
+      1) if [ "$exit" -eq 0 ]; then exit=1; fi ;;
+      2) exit=2 ;;
+      *) return "$file_rc" ;;
+    esac
+  done
+  return "$exit"
+}
+
+main "$@"
diff --git a/.github/workflows/feature-ideation-tests.yml b/.github/workflows/feature-ideation-tests.yml
index eeac386b0..9bb808286 100644
--- a/.github/workflows/feature-ideation-tests.yml
+++ b/.github/workflows/feature-ideation-tests.yml
@@ -61,8 +61,8 @@ jobs:
           sudo apt-get update -qq
           sudo apt-get install -y --no-install-recommends bats shellcheck jq
 
-      - name: Install Python jsonschema
-        run: pip install --quiet 'jsonschema>=4'
+      - name: Install Python jsonschema and PyYAML
+        run: pip install --quiet 'jsonschema>=4' 'pyyaml>=6'
 
       - name: shellcheck
         run: |
@@ -71,6 +71,7 @@ jobs:
           shellcheck -x \
             collect-signals.sh \
             lint-prompt.sh \
+            lint-caller.sh \
             match-discussions.sh \
             discussion-mutations.sh \
             lib/gh-safe.sh \
@@ -81,6 +82,9 @@ jobs:
       - name: Lint direct prompt blocks
         run: bash .github/scripts/feature-ideation/lint-prompt.sh
 
+      - name: Lint caller stubs for the #571 inputs-context startup failure
+        run: bash .github/scripts/feature-ideation/lint-caller.sh
+
       - name: Validate schema fixtures
         run: |
           set -euo pipefail
diff --git a/.github/workflows/feature-ideation.yml b/.github/workflows/feature-ideation.yml
index 96d80705d..0d7e5426c 100644
--- a/.github/workflows/feature-ideation.yml
+++ b/.github/workflows/feature-ideation.yml
@@ -1,19 +1,44 @@
 # ─────────────────────────────────────────────────────────────────────────────
 # SOURCE OF TRUTH: petry-projects/.github/standards/workflows/feature-ideation.yml
-# Standard:        petry-projects/.github/standards/ci-standards.md#8
+# Standard:        petry-projects/.github/standards/ci-standards.md#8-feature-ideation-feature-ideationyml--bmad-method-repos
 # Reusable:        petry-projects/.github/.github/workflows/feature-ideation-reusable.yml
 #
 # AGENTS — READ BEFORE EDITING:
 #   • This file is a THIN CALLER STUB. The 5-phase ideation pipeline, the
 #     Opus 4.6 model selection, the github_token override, and the
 #     ANTHROPIC_MODEL env var all live in the reusable workflow above.
-#   • You MAY change: the `project_context` value and optionally the cron
-#     schedule.
+#   • You MAY change: the `project_context` value (the only required edit
+#     per repo), and optionally the cron schedule.
 #   • You MUST NOT change: trigger event shape, the `uses:` line, the
-#     job-level `permissions:` block, or the `secrets:` block.
+#     job-level `permissions:` block, or the `secrets:` block — these are
+#     required for the reusable to work.
 #   • If you need different behaviour, open a PR against the reusable in
-#     this repo. The change will propagate after the v1 tag is bumped.
+#     the central repo. The change will propagate everywhere on next run.
 # ─────────────────────────────────────────────────────────────────────────────
+#
+# Feature Ideation workflow stub — for BMAD Method-enabled repos.
+#
+# This is a thin caller for the org-wide reusable workflow at
+# petry-projects/.github/.github/workflows/feature-ideation-reusable.yml
+# All ideation logic, the multi-skill pipeline, the Opus 4.6 model
+# selection, and the github_token override live in the reusable workflow.
+#
+# To adopt:
+#   1. Copy this file to .github/workflows/feature-ideation.yml in your repo.
+#   2. Replace the `project_context` value with a 3-5 sentence description
+#      of your project, its target users, and the competitive landscape Mary
+#      should research. This is the only required customisation.
+#   3. (Optional) Copy standards/feature-ideation-sources.md from
+#      petry-projects/.github to .github/feature-ideation-sources.md in your
+#      repo and trim/extend it for your project. Mary uses YOUR copy — not the
+#      central template — so each repo controls its own source list.
+#      Pass `sources_file: path/to/your-list.md` to the reusable workflow if
+#      you prefer a different location.
+#   4. (Optional) Adjust the schedule cron if Friday morning UTC doesn't suit.
+#   5. Ensure GitHub Discussions is enabled with an "Ideas" category.
+#   6. Confirm the org-level secret CLAUDE_CODE_OAUTH_TOKEN is accessible.
+#
+# Standard: https://github.com/petry-projects/.github/blob/main/standards/ci-standards.md#8-feature-ideation-feature-ideationyml--bmad-method-repos
 name: Feature Research & Ideation (BMAD Analyst)
 
 on:
@@ -22,7 +47,7 @@ on:
   workflow_dispatch:
     inputs:
       focus_area:
-        description: 'Optional focus area (e.g., "CI patterns", "agent security")'
+        description: 'Optional focus area (e.g., "accessibility", "performance")'
         required: false
         type: string
       research_depth:
@@ -35,12 +60,23 @@ on:
           - standard
           - deep
       dry_run:
-        description: 'Log Discussion mutations to artifact instead of writing live'
+        description: 'Skip Discussion mutations and log them to a JSONL artifact instead. Use this on a fork to smoke-test before going live.'
         required: false
         default: false
         type: boolean
+      enhance_backlog:
+        description: 'Backlog enhancement sweep: enhance this repo''s open, human-authored, not-yet-enhanced Ideas (one comment each, idempotent). Combine with dry_run=true to preview.'
+        required: false
+        default: false
+        type: boolean
+      target_discussion:
+        description: 'Internal — set by `redispatch` when bridging a `discussion:created` event to workflow_dispatch (claude-code-action does not support discussion contexts). Not for manual use.'
+        required: false
+        default: ''
+        type: string
   # Auto-enhance a freshly-created idea: when a new Discussion is opened in the
-  # Ideas category, the analyst researches and refines it in single-idea mode.
+  # Ideas category, the `redispatch` job bridges it to workflow_dispatch (see below)
+  # so the analyst refines it in single-idea mode under a supported event.
   discussion:
     types: [created]
 
@@ -51,36 +87,109 @@ concurrency:
   cancel-in-progress: false
 
 jobs:
-  ideate:
-    # On the `discussion` trigger, only run for new ideas in the Ideas category
-    # and skip the bot's own creations (avoids re-enhancing scheduled output;
-    # enhancement posts a comment, which does not re-fire `discussion: created`).
-    # Schedule/dispatch runs are unaffected by this guard.
+  # ── #963 discussion→dispatch redispatch bridge ───────────────────────────────
+  # claude-code-action aborts on `discussion` event contexts ("Unsupported event
+  # type: discussion"). So on `discussion: created` we do NOT call the reusable
+  # inline; the `redispatch` job re-invokes this workflow via workflow_dispatch
+  # (under which the action runs cleanly), passing the number as target_discussion.
+  # Mirrors initiative-planner.yml's redispatch (#618). The `ideate` reusable call
+  # then runs only on workflow_dispatch/schedule.
+  redispatch:
     if: >-
-      github.event_name != 'discussion' ||
-      (github.event.discussion.category.slug == 'ideas' &&
-       github.event.discussion.user.type != 'Bot')
+      github.event_name == 'discussion' &&
+      github.event.discussion.category.slug == 'ideas' &&
+      github.event.discussion.user.type != 'Bot'
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    permissions: {}
+    steps:
+      - name: Guard — PAT present
+        # A workflow_dispatch fired with GITHUB_TOKEN is accepted but never starts
+        # a run (loop prevention), so a PAT is required (same as initiative-planner).
+        env:
+          GH_TOKEN: ${{ secrets.GH_PAT_WORKFLOWS }}
+        run: |
+          if [ -z "${GH_TOKEN}" ]; then
+            echo "::error::GH_PAT_WORKFLOWS is required — a workflow_dispatch fired with GITHUB_TOKEN will not start a run."
+            exit 1
+          fi
+      - name: Re-dispatch under workflow_dispatch
+        env:
+          GH_TOKEN: ${{ secrets.GH_PAT_WORKFLOWS }}
+          REPO: ${{ github.repository }}
+          DISCUSSION_NUMBER: ${{ github.event.discussion.number }}
+        run: |
+          gh workflow run feature-ideation.yml --repo "${REPO}" \
+            -f target_discussion="${DISCUSSION_NUMBER}"
+
+  # ── #571 input resolution — keep `inputs` out of the reusable `with:` ────────
+  # A reusable-workflow call graph (its `uses:` + `with:`) is validated at
+  # WORKFLOW SETUP, before — and regardless of — the calling job's `if:`. The
+  # `inputs` context is only populated for workflow_dispatch/workflow_call, so a
+  # `with:` that references `${{ inputs.* }}` trips a zero-job startup failure on
+  # the `discussion` event even though `ideate` is gated off it. We therefore
+  # resolve the dispatch inputs in this ordinary job (whose step expressions are
+  # evaluated at RUN time and are skipped on `discussion`) and hand them to
+  # `ideate` via `needs.prep.outputs.*` — an always-valid context that defers the
+  # `with:` evaluation to run time. Do not reference `inputs` in `ideate.with`.
+  prep:
+    if: github.event_name != 'discussion'
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    permissions: {}
+    outputs:
+      target_discussion: ${{ steps.resolve.outputs.target_discussion }}
+      focus_area: ${{ steps.resolve.outputs.focus_area }}
+      research_depth: ${{ steps.resolve.outputs.research_depth }}
+      dry_run: ${{ steps.resolve.outputs.dry_run }}
+      enhance_backlog: ${{ steps.resolve.outputs.enhance_backlog }}
+    steps:
+      - name: Resolve dispatch inputs
+        id: resolve
+        # `github.event.inputs.*` (not the `inputs` context) so this reads
+        # cleanly on schedule too, where it is simply null → the defaults below.
+        env:
+          TARGET_DISCUSSION: ${{ github.event.inputs.target_discussion }}
+          FOCUS_AREA: ${{ github.event.inputs.focus_area }}
+          RESEARCH_DEPTH: ${{ github.event.inputs.research_depth }}
+          DRY_RUN: ${{ github.event.inputs.dry_run }}
+          ENHANCE_BACKLOG: ${{ github.event.inputs.enhance_backlog }}
+        run: |
+          {
+            echo "target_discussion=${TARGET_DISCUSSION:-}"
+            echo "focus_area=${FOCUS_AREA:-}"
+            echo "research_depth=${RESEARCH_DEPTH:-standard}"
+            echo "dry_run=${DRY_RUN:-false}"
+            echo "enhance_backlog=${ENHANCE_BACKLOG:-false}"
+          } >> "$GITHUB_OUTPUT"
+
+  ideate:
+    # Runs on workflow_dispatch (including the redispatch bridge above) and schedule.
+    needs: [prep]
+    if: github.event_name != 'discussion'
     # Permissions cascade from the calling job to the reusable workflow.
     # The reusable workflow's two jobs (gather-signals + analyze) need:
     #   - contents:      read   (checkout, file reads)
     #   - issues:        read   (signal collection)
     #   - pull-requests: read   (signal collection)
     #   - discussions:   write  (CRITICAL — create/update Discussion threads)
-    #   - actions:       read   (feed checkpoint via gh run list)
     #   - id-token:      write  (claude-code-action OIDC for GitHub App token)
+    #   - actions:       read   (feed checkpoint — last successful run query)
     permissions:
       contents: read
       issues: read
       pull-requests: read
       discussions: write
-      actions: read
       id-token: write
+      actions: read
     uses: petry-projects/.github/.github/workflows/feature-ideation-reusable.yml@3500ed2675d2ac60c8521f295ffbeba75434b953 # v1
     with:
-      # On the `discussion: created` trigger this carries the new Discussion's
-      # number → the reusable runs single-idea enhancement mode. Empty on
-      # schedule/dispatch → normal scan.
-      target_discussion: ${{ github.event.discussion.number }}
+      # All values below come from `needs.prep.outputs.*` (resolved in the `prep`
+      # job) — NOT the `inputs` context — so this reusable `with:` compiles on the
+      # `discussion` event without a zero-job startup failure (#571).
+      # `target_discussion` is forwarded by the `redispatch` bridge above on
+      # `discussion: created`; empty on schedule/dispatch → normal scan.
+      target_discussion: ${{ needs.prep.outputs.target_discussion }}
       project_context: |
         petry-projects/.github is the org-level standards and tooling repository
         for the petry-projects GitHub organisation. It owns the canonical CI/CD
@@ -97,8 +206,12 @@ jobs:
         scopes), and developer-experience standards for LLM-heavy codebases.
         No public users — this is internal DevX infrastructure.
       sources_file: 'standards/feature-ideation-sources.md'
-      focus_area: ${{ inputs.focus_area || '' }}
-      research_depth: ${{ inputs.research_depth || 'standard' }}
-      dry_run: ${{ inputs.dry_run || false }}
+      focus_area: ${{ needs.prep.outputs.focus_area }}
+      research_depth: ${{ needs.prep.outputs.research_depth }}
+      # prep emits the string 'true'/'false'; fromJSON casts it to the boolean
+      # the reusable's typed inputs require.
+      dry_run: ${{ fromJSON(needs.prep.outputs.dry_run) }}
+      # Backlog enhancement sweep of existing un-enhanced Ideas (workflow_dispatch).
+      enhance_backlog: ${{ fromJSON(needs.prep.outputs.enhance_backlog) }}
     secrets:
       CLAUDE_CODE_OAUTH_TOKEN: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
diff --git a/standards/ci-standards.md b/standards/ci-standards.md
index 7d1a33ea3..383716bae 100644
--- a/standards/ci-standards.md
+++ b/standards/ci-standards.md
@@ -963,13 +963,29 @@ is a separate Discussion, updated by subsequent runs as the market and project
 evolve.
 
 **Triggers:** the weekly `schedule`, manual `workflow_dispatch`, **and
-`discussion: created`**. On the discussion trigger, the stub passes
-`target_discussion: ${{ github.event.discussion.number }}`, putting the reusable
-in **single-idea enhancement mode**: it researches and refines that one new idea
-and posts a single enhancement comment, rather than running the broad scan. A
-job-level `if` restricts this to new Discussions in the **Ideas** category and
-skips the bot's own creations; enhancement is a comment (which does not re-fire
-`created`), so there is no trigger loop.
+`discussion: created`**. `claude-code-action` aborts on `discussion` event
+contexts, so the stub does **not** call the reusable inline on that event.
+Instead a `redispatch` job — gated to new Discussions in the **Ideas** category,
+skipping the bot's own creations — re-invokes the workflow via
+`workflow_dispatch` (using `GH_PAT_WORKFLOWS`), forwarding the Discussion number
+as the `target_discussion` input. The re-dispatched run puts the reusable in
+**single-idea enhancement mode**: it researches and refines that one new idea and
+posts a single enhancement comment, rather than running the broad scan.
+Enhancement is a comment (which does not re-fire `created`), so there is no
+trigger loop. This mirrors `initiative-planner.yml`'s redispatch bridge.
+
+> **#571 — never reference the `inputs` context in the reusable `with:`.** A
+> reusable-workflow call graph (`uses:` + `with:`) is validated at **workflow
+> setup**, before and regardless of the calling job's `if:`. The `inputs`
+> context is only populated for `workflow_dispatch` / `workflow_call`, so a
+> `with:` value referencing `${{ inputs.* }}` fails the whole run (zero jobs,
+> "Invalid workflow file") on the `discussion` trigger even though the `ideate`
+> job is gated off it. The stub therefore resolves dispatch inputs in an
+> ordinary `prep` job (whose step expressions run at job time and are skipped on
+> `discussion`) and passes them to `ideate` via `needs.prep.outputs.*` — an
+> always-valid context that defers the `with:` evaluation to run time. This is
+> enforced by [`lint-caller.sh`](../.github/scripts/feature-ideation/lint-caller.sh)
+> in the feature-ideation test suite.
 
 **Backlog enhancement (backfill) + dry-run.** Beyond enhancing *newly-created*
 Ideas, the reusable can **backfill the existing Ideas backlog**: dispatch with
diff --git a/standards/workflows/feature-ideation.yml b/standards/workflows/feature-ideation.yml
index c7ed75fea..f5ccbba78 100644
--- a/standards/workflows/feature-ideation.yml
+++ b/standards/workflows/feature-ideation.yml
@@ -122,8 +122,50 @@ jobs:
           gh workflow run feature-ideation.yml --repo "${REPO}" \
             -f target_discussion="${DISCUSSION_NUMBER}"
 
+  # ── #571 input resolution — keep `inputs` out of the reusable `with:` ────────
+  # A reusable-workflow call graph (its `uses:` + `with:`) is validated at
+  # WORKFLOW SETUP, before — and regardless of — the calling job's `if:`. The
+  # `inputs` context is only populated for workflow_dispatch/workflow_call, so a
+  # `with:` that references `${{ inputs.* }}` trips a zero-job startup failure on
+  # the `discussion` event even though `ideate` is gated off it. We therefore
+  # resolve the dispatch inputs in this ordinary job (whose step expressions are
+  # evaluated at RUN time and are skipped on `discussion`) and hand them to
+  # `ideate` via `needs.prep.outputs.*` — an always-valid context that defers the
+  # `with:` evaluation to run time. Do not reference `inputs` in `ideate.with`.
+  prep:
+    if: github.event_name != 'discussion'
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    permissions: {}
+    outputs:
+      target_discussion: ${{ steps.resolve.outputs.target_discussion }}
+      focus_area: ${{ steps.resolve.outputs.focus_area }}
+      research_depth: ${{ steps.resolve.outputs.research_depth }}
+      dry_run: ${{ steps.resolve.outputs.dry_run }}
+      enhance_backlog: ${{ steps.resolve.outputs.enhance_backlog }}
+    steps:
+      - name: Resolve dispatch inputs
+        id: resolve
+        # `github.event.inputs.*` (not the `inputs` context) so this reads
+        # cleanly on schedule too, where it is simply null → the defaults below.
+        env:
+          TARGET_DISCUSSION: ${{ github.event.inputs.target_discussion }}
+          FOCUS_AREA: ${{ github.event.inputs.focus_area }}
+          RESEARCH_DEPTH: ${{ github.event.inputs.research_depth }}
+          DRY_RUN: ${{ github.event.inputs.dry_run }}
+          ENHANCE_BACKLOG: ${{ github.event.inputs.enhance_backlog }}
+        run: |
+          {
+            echo "target_discussion=${TARGET_DISCUSSION:-}"
+            echo "focus_area=${FOCUS_AREA:-}"
+            echo "research_depth=${RESEARCH_DEPTH:-standard}"
+            echo "dry_run=${DRY_RUN:-false}"
+            echo "enhance_backlog=${ENHANCE_BACKLOG:-false}"
+          } >> "$GITHUB_OUTPUT"
+
   ideate:
     # Runs on workflow_dispatch (including the redispatch bridge above) and schedule.
+    needs: [prep]
     if: github.event_name != 'discussion'
     # Permissions cascade from the calling job to the reusable workflow.
     # The reusable workflow's two jobs (gather-signals + analyze) need:
@@ -142,10 +184,12 @@ jobs:
       actions: read
     uses: petry-projects/.github/.github/workflows/feature-ideation-reusable.yml@e20a8fac1b6a10bd6ad0999258e88a423f890ba6 # v1
     with:
-      # Populated by the `redispatch` bridge above (which forwards the new
-      # Discussion number on `discussion: created`); empty on schedule/dispatch →
-      # normal scan. Do not edit this line.
-      target_discussion: ${{ inputs.target_discussion || '' }}
+      # All values below come from `needs.prep.outputs.*` (resolved in the `prep`
+      # job) — NOT the `inputs` context — so this reusable `with:` compiles on the
+      # `discussion` event without a zero-job startup failure (#571).
+      # `target_discussion` is forwarded by the `redispatch` bridge above on
+      # `discussion: created`; empty on schedule/dispatch → normal scan.
+      target_discussion: ${{ needs.prep.outputs.target_discussion }}
       # === CUSTOMISE THIS PER REPO — the only required edit ===
       # Replace this paragraph with a 3-5 sentence description of your project,
       # its target users, and the competitive landscape. The more specific you
@@ -160,10 +204,12 @@ jobs:
       # workflow defaults to that path, so you only need to uncomment and change
       # sources_file below if you store the list somewhere else.
       # sources_file: 'docs/feature-ideation-sources.md'
-      focus_area: ${{ inputs.focus_area || '' }}
-      research_depth: ${{ inputs.research_depth || 'standard' }}
-      dry_run: ${{ inputs.dry_run || false }}
+      focus_area: ${{ needs.prep.outputs.focus_area }}
+      research_depth: ${{ needs.prep.outputs.research_depth }}
+      # prep emits the string 'true'/'false'; fromJSON casts it to the boolean
+      # the reusable's typed inputs require.
+      dry_run: ${{ fromJSON(needs.prep.outputs.dry_run) }}
       # Backlog enhancement sweep of existing un-enhanced Ideas (workflow_dispatch).
-      enhance_backlog: ${{ inputs.enhance_backlog == true }}
+      enhance_backlog: ${{ fromJSON(needs.prep.outputs.enhance_backlog) }}
     secrets:
       CLAUDE_CODE_OAUTH_TOKEN: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}
diff --git a/test/workflows/feature-ideation/fixtures/callers/clean-needs-outputs.yml b/test/workflows/feature-ideation/fixtures/callers/clean-needs-outputs.yml
new file mode 100644
index 000000000..0d9f5d548
--- /dev/null
+++ b/test/workflows/feature-ideation/fixtures/callers/clean-needs-outputs.yml
@@ -0,0 +1,35 @@
+# Fixture: the #571-correct shape — dispatch inputs resolved in a `prep` job and
+# passed to the reusable via `needs.prep.outputs.*`. No `inputs` context in the
+# reusable `with:`. lint-caller.sh must pass this (exit 0).
+name: clean caller
+on:
+  schedule:
+    - cron: '0 7 * * 5'
+  workflow_dispatch:
+    inputs:
+      focus_area:
+        required: false
+        type: string
+  discussion:
+    types: [created]
+
+jobs:
+  prep:
+    if: github.event_name != 'discussion'
+    runs-on: ubuntu-latest
+    outputs:
+      focus_area: ${{ steps.resolve.outputs.focus_area }}
+    steps:
+      - id: resolve
+        env:
+          FOCUS_AREA: ${{ github.event.inputs.focus_area }}
+        run: echo "focus_area=${FOCUS_AREA:-}" >> "$GITHUB_OUTPUT"
+
+  ideate:
+    needs: [prep]
+    if: github.event_name != 'discussion'
+    uses: org/.github/.github/workflows/reusable.yml@v1
+    with:
+      project_context: |
+        A literal string with no expressions.
+      focus_area: ${{ needs.prep.outputs.focus_area }}
diff --git a/test/workflows/feature-ideation/fixtures/callers/inputs-in-with.yml b/test/workflows/feature-ideation/fixtures/callers/inputs-in-with.yml
new file mode 100644
index 000000000..82882c46e
--- /dev/null
+++ b/test/workflows/feature-ideation/fixtures/callers/inputs-in-with.yml
@@ -0,0 +1,24 @@
+# Fixture: the #571-broken shape — the reusable `with:` references the `inputs`
+# context. The job `if:` does NOT prevent the reusable call graph from being
+# validated at workflow setup, so on the `discussion` trigger this fails the
+# whole run with zero jobs. lint-caller.sh must flag this (exit 1).
+name: broken caller
+on:
+  workflow_dispatch:
+    inputs:
+      focus_area:
+        required: false
+        type: string
+      target_discussion:
+        required: false
+        type: string
+  discussion:
+    types: [created]
+
+jobs:
+  ideate:
+    if: github.event_name != 'discussion'
+    uses: org/.github/.github/workflows/reusable.yml@v1
+    with:
+      target_discussion: ${{ inputs.target_discussion || '' }}
+      focus_area: ${{ inputs.focus_area || '' }}
diff --git a/test/workflows/feature-ideation/lint-caller.bats b/test/workflows/feature-ideation/lint-caller.bats
new file mode 100644
index 000000000..f9f6f8200
--- /dev/null
+++ b/test/workflows/feature-ideation/lint-caller.bats
@@ -0,0 +1,127 @@
+#!/usr/bin/env bats
+# Tests for lint-caller.sh — kills #571 (a reusable-workflow `with:` that
+# references the `inputs` context, which fails at workflow setup with zero jobs
+# on the `discussion` trigger regardless of the calling job's `if:`).
+
+load 'helpers/setup'
+
+setup() {
+  tt_make_tmpdir
+}
+
+teardown() {
+  tt_cleanup_tmpdir
+}
+
+LINTER="${TT_REPO_ROOT}/.github/scripts/feature-ideation/lint-caller.sh"
+CALLERS="${TT_FIXTURES_DIR}/callers"
+
+write_yml() {
+  local path="$1"
+  cat >"$path"
+}
+
+# ---------------------------------------------------------------------------
+# Committed fixtures — the correct and the broken shapes.
+# ---------------------------------------------------------------------------
+
+@test "lint-caller: clean needs.prep.outputs caller passes" {
+  run bash "$LINTER" "${CALLERS}/clean-needs-outputs.yml"
+  [ "$status" -eq 0 ]
+}
+
+@test "lint-caller: FAILS on inputs.* in a reusable with: block" {
+  run bash "$LINTER" "${CALLERS}/inputs-in-with.yml"
+  [ "$status" -eq 1 ]
+  # The offending key must be named in the output.
+  [[ "$output" == *"with.target_discussion"* ]] || [[ "$output" == *"with.focus_area"* ]]
+}
+
+# ---------------------------------------------------------------------------
+# False-positive guards.
+# ---------------------------------------------------------------------------
+
+@test "lint-caller: step-level action uses: with inputs is NOT flagged" {
+  write_yml "${TT_TMP}/steplevel.yml" <<'YML'
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/github-script@v7
+        with:
+          script: core.info('${{ inputs.focus_area }}')
+YML
+  run bash "$LINTER" "${TT_TMP}/steplevel.yml"
+  [ "$status" -eq 0 ]
+}
+
+@test "lint-caller: github.event.inputs.* in a reusable with: is allowed" {
+  # github.event.inputs is available at setup for every event (null when absent),
+  # so it does not cause the #571 startup failure and must not be flagged.
+  write_yml "${TT_TMP}/ghevent.yml" <<'YML'
+jobs:
+  ideate:
+    uses: org/.github/.github/workflows/reusable.yml@v1
+    with:
+      focus_area: ${{ github.event.inputs.focus_area }}
+YML
+  run bash "$LINTER" "${TT_TMP}/ghevent.yml"
+  [ "$status" -eq 0 ]
+}
+
+@test "lint-caller: a job without a reusable uses: is ignored" {
+  write_yml "${TT_TMP}/nouses.yml" <<'YML'
+jobs:
+  prep:
+    runs-on: ubuntu-latest
+    steps:
+      - run: echo "${{ inputs.focus_area }}"
+YML
+  run bash "$LINTER" "${TT_TMP}/nouses.yml"
+  [ "$status" -eq 0 ]
+}
+
+@test "lint-caller: inputs referenced via index syntax is flagged" {
+  write_yml "${TT_TMP}/index.yml" <<'YML'
+jobs:
+  ideate:
+    uses: org/.github/.github/workflows/reusable.yml@v1
+    with:
+      focus_area: ${{ inputs['focus_area'] }}
+YML
+  run bash "$LINTER" "${TT_TMP}/index.yml"
+  [ "$status" -eq 1 ]
+}
+
+# ---------------------------------------------------------------------------
+# Regression guards — the live caller stubs must stay clean.
+# ---------------------------------------------------------------------------
+
+@test "lint-caller: standards caller-stub template lints clean" {
+  workflow="${TT_REPO_ROOT}/standards/workflows/feature-ideation.yml"
+  [ -f "$workflow" ]
+  run bash "$LINTER" "$workflow"
+  [ "$status" -eq 0 ]
+}
+
+@test "lint-caller: this repo's own caller lints clean" {
+  workflow="${TT_REPO_ROOT}/.github/workflows/feature-ideation.yml"
+  [ -f "$workflow" ]
+  run bash "$LINTER" "$workflow"
+  [ "$status" -eq 0 ]
+}
+
+@test "lint-caller: default (no args) scans both caller stubs and passes" {
+  run bash "$LINTER"
+  [ "$status" -eq 0 ]
+}
+
+# ---------------------------------------------------------------------------
+# Exit-code precedence: a file/parse error (2) must not be downgraded by a
+# later lint finding (1). Mirrors lint-prompt.sh.
+# ---------------------------------------------------------------------------
+
+@test "lint-caller: missing file before a lint-failing file still exits 2" {
+  run bash "$LINTER" "${TT_TMP}/missing.yml" "${CALLERS}/inputs-in-with.yml"
+  [ "$status" -eq 2 ]
+}