prizmkit 1.1.109 → 1.1.111

package/bundled/VERSION.json

@@ -1,5 +1,5 @@
 {
-  "frameworkVersion": "1.1.109",
-  "bundledAt": "2026-07-08T05:29:47.417Z",
-  "bundledFrom": "814bf2f"
+  "frameworkVersion": "1.1.111",
+  "bundledAt": "2026-07-08T07:06:52.766Z",
+  "bundledFrom": "1db214b"
 }

package/bundled/agents/prizm-dev-team-reviewer.md

@@ -25,6 +25,8 @@ You are the team's "senior engineer doing code review" — you diagnose problems
 2. If no `context-snapshot.md`: read `spec.md`, `plan.md` from the artifact directory, then `.prizmkit/prizm-docs/root.prizm` and relevant L1/L2 docs
 3. Identify changed files from `## Implementation Log` or completed tasks in plan.md
 
+**Worktree constraint**: You are spawned with default/no worktree isolation by the orchestrator. You review the active checkout's changes — the orchestrator provides workspace context (git status, diff summary, new-file paths). Do NOT request or create `.claude/worktrees/...` tool worktrees. Do NOT re-discover or re-index the whole repository; use the orchestrator-provided context as your review scope and request only targeted file reads.
+
 **File Reading Rule**: Read ONLY files listed in the Implementation Log for diagnosis — do not explore unrelated files. Exception: during Fix Strategy Formulation, you MAY read additional files to trace impact (callers, dependents, shared patterns).
 
 ### Review Workflow
@@ -74,6 +76,8 @@ When Dev has applied fixes and returns for re-review:
 - Do not modify source files to fix issues — produce Fix Instructions for Dev instead
 - **Do NOT run test suites** — the Orchestrator handles testing. You review code, not execute tests.
 - Do NOT re-read source files already listed in context-snapshot.md Section 4 File Manifest unless you need a specific code detail for a finding
+- **Do NOT request or create worktrees** (`.claude/worktrees/...` or similar) — you review the active checkout's changes using the orchestrator-provided workspace context
+- **Do NOT re-discover or broadly re-index the repository** — the orchestrator has already captured the diff scope; request only targeted file reads for lines you need
 
 ### Behavioral Rules
 

package/bundled/dev-pipeline/scripts/generate-bootstrap-prompt.py

@@ -190,9 +190,42 @@ def detect_test_commands(project_root):
     """
     test_commands = []
 
-    # Check for npm/package.json
-    if os.path.exists(os.path.join(project_root, "package.json")):
-        test_commands.append("npm test")
+    # Check for npm/package.json — use content-aware detection
+    pkg_json_path = os.path.join(project_root, "package.json")
+    if os.path.exists(pkg_json_path):
+        try:
+            with open(pkg_json_path, "r", encoding="utf-8") as f:
+                pkg_data = json.load(f)
+            scripts = pkg_data.get("scripts", {}) if isinstance(pkg_data, dict) else {}
+
+            # Detect package manager from lockfiles
+            if os.path.exists(os.path.join(project_root, "pnpm-lock.yaml")):
+                pm_cmd = "pnpm"
+            elif os.path.exists(os.path.join(project_root, "yarn.lock")):
+                pm_cmd = "yarn"
+            else:
+                pm_cmd = "npm"
+
+            # Find best test script: prefer "test", then "test:unit",
+            # then first script matching test-related patterns
+            if "test" in scripts:
+                test_commands.append("{} test".format(pm_cmd))
+            elif "test:unit" in scripts:
+                test_commands.append("{} run test:unit".format(pm_cmd))
+            else:
+                # Search for any test-like script name
+                test_like = None
+                for script_name in sorted(scripts.keys()):
+                    if "test" in script_name.lower():
+                        test_like = script_name
+                        break
+                if test_like:
+                    test_commands.append(
+                        "{} run {}".format(pm_cmd, test_like)
+                    )
+        except Exception:
+            # Corrupt/unreadable package.json → skip JS detection
+            pass
 
     # Check for Go
     if os.path.exists(os.path.join(project_root, "go.mod")):

package/bundled/dev-pipeline/templates/bootstrap-tier1.md

@@ -46,15 +46,15 @@ You are running in **headless non-interactive mode** with a FINITE context windo
 
 1. **context-snapshot.md is your single source of truth** — After Phase 1 builds it, ALWAYS read context-snapshot.md instead of re-reading individual source files
 2. **Never re-read your own writes** — After you create/modify a file, do NOT read it back to verify. Trust your write was correct.
-3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use `grep -n '<pattern>' <file>` or `sed -n '<start>,<end>p' <file>` for targeted line lookups instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
+3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use platform file/search tools for targeted lookup; when shell-free helper lookup is needed, run `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<pattern>" --json`, then read only the smallest bounded range with the platform file-reading tool instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
 4. **Stay focused** — Do NOT explore code unrelated to this feature. No curiosity-driven reads.
 4. **One task at a time** — In Phase 3 (implement), complete and test one task before starting the next.
-5. **Minimize tool output** — Never load full command output into context. First capture to a temp file (`cmd 2>&1 | tee /tmp/out.txt | tail -20`), then scan the head/tail to identify relevant fields, and use targeted filtering (`grep`, `sed`, `awk`) to extract only the information needed for the current task. Only read the filtered result — never the raw full output.
+5. **Minimize tool output** — Never load full command output into context. First capture command output to a temp file using the command/tool output option or the host-appropriate redirection, then inspect only the final relevant lines and use targeted search helpers to extract the information needed for the current task. Only read the filtered result — never the raw full output.
 6. **No intermediate commits** — Do NOT run `git add`/`git commit` during Phase 1-3. All changes are committed once at the end in Phase 4 via `/prizmkit-committer`.
-7. **Capture test output once** — When running test suites, always use `<test-command> 2>&1 | tee /tmp/test-out.txt | tail -20`. Then grep `/tmp/test-out.txt` for details. Never re-run the suite just to apply a different filter.
+7. **Capture test output once** — When running test suites, capture output once to `/tmp/test-out.txt` using the command/tool output option or the host-appropriate redirection, surface only the final relevant lines, then search `/tmp/test-out.txt` for details. Never re-run the suite just to apply a different filter.
 9. **Package version verification (BLOCKING)** — Before writing ANY dependency version, verify it exists via registry query (`npm view <pkg> dist-tags.latest`, `pip index versions <pkg>`, etc.). NEVER guess version numbers.
 10. **Build artifact hygiene** — After build/compile, ensure output is in `.gitignore`. Never commit binaries/build output.
-11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Run `wc -l <file>` to get real line count, then `sed -n` for exact range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets — `grep -n '<anchor>'` first.
+11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Use the platform file-reading tool to refresh bounded line information, or locate an anchor with `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<anchor>" --json` before reading a small bounded range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets or prescribe POSIX-only lookup commands.
 
 ---
 
@@ -95,9 +95,9 @@ If MISSING — build it now:
 3. Scan the detected source directories for files related to this feature; read each one
 4. Write `.prizmkit/specs/{{FEATURE_SLUG}}/context-snapshot.md`:
    - **Section 1 — Feature Brief**: feature description + acceptance criteria (copy from above)
-   - **Section 2 — Project Structure**: run the following to get a visual directory tree, then paste output:
+   - **Section 2 — Project Structure**: run the helper to get a visual directory tree, then paste output:
      ```bash
-     find . -maxdepth 2 -type d -not -path '*/node_modules/*' -not -path '*/.git/*' -not -path '*/dist/*' -not -path '*/build/*' -not -path '*/__pycache__/*' -not -path '*/vendor/*' | sed -e 's;[^/]*/;|____;g;s;____|; |;g'
+     {{RUNTIME_HELPER_CMD}} artifact tree . --max-depth 2
      ```
    - **Section 3 — Prizm Context**: content of root.prizm and relevant L1/L2 docs
    - **Section 4 — Existing Source Files**: **full verbatim content** of each related file in fenced code blocks (with `### path/to/file` heading and line count). Include ALL files needed for implementation and review — downstream phases read this section instead of re-reading individual source files
@@ -126,9 +126,9 @@ Before proceeding past CP-1:
 **Build artifacts**: After any build/compile command (`go build`, `npm run build`, `tsc`, etc.), ensure the output binary or build directory is in `.gitignore`:
 ```bash
 # Example for Go
-grep -q '^/binary-name$' .gitignore || echo '/binary-name' >> .gitignore
+{{RUNTIME_HELPER_CMD}} text contains .gitignore --pattern "/binary-name" --found-marker IGNORED --missing-marker NEEDS_IGNORE_ENTRY
 ```
-Never commit compiled binaries, build output, or generated artifacts.
+If the helper prints `NEEDS_IGNORE_ENTRY`, edit `.gitignore` and add `/binary-name` as a single line. Never commit compiled binaries, build output, or generated artifacts.
 
 **3a.** Run `/prizmkit-implement` — this handles the implementation cycle:
 - Reads plan.md Tasks section from `.prizmkit/specs/{{FEATURE_SLUG}}/`
@@ -226,7 +226,7 @@ playwright-cli --help
 ```bash
 {{RUNTIME_HELPER_CMD}} platform skill playwright-cli --prefix
 ```
-If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy installed skill from `$HOME/.claude/skills/playwright-cli` to the path shown by the helper output for the current platform.
+If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy the installed skill from the Claude user skills directory to the path shown by the helper output for the current platform.
 
 0d. Read the installed playwright-cli skill (SKILL.md) for workflow guidance. Use its recommended patterns to construct your verification flow.
 
@@ -238,14 +238,14 @@ You know this project's tech stack. Detect and start the dev server yourself:
 2. Determine the dev server port from the pre-detected pipeline value (`{{DEV_PORT}}`) or project config. Do NOT hardcode or guess the port. Record it as `DEV_PORT` in your notes.
 3. Verify the port is available:
    ```bash
-   {{RUNTIME_HELPER_CMD}} port check $DEV_PORT
+   {{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}
    ```
 4. Start the dev server in background, capture PID:
    ```bash
    {{RUNTIME_HELPER_CMD}} process start --pid-file .prizmkit/state/browser-dev-server.pid -- <start-command tokens>
    ```
-5. Wait for server to be ready: run `{{RUNTIME_HELPER_CMD}} web wait-ready --url http://localhost:$DEV_PORT --timeout 30 --interval 2`
-6. Open the app in playwright-cli: `playwright-cli open http://localhost:$DEV_PORT`
+5. Wait for server to be ready: run `{{RUNTIME_HELPER_CMD}} web wait-ready --url http://localhost:{{DEV_PORT}} --timeout 30 --interval 2`
+6. Open the app in playwright-cli: `playwright-cli open http://localhost:{{DEV_PORT}}`
 7. If the page requires authentication, use playwright-cli to register a test user and log in first
 
 **Step 2 — Verification**:
@@ -259,7 +259,7 @@ Construct your verification workflow based on: (1) the playwright-cli skill docu
 
 1. Close the playwright-cli browser: `playwright-cli close`
 2. Kill the dev server process: `{{RUNTIME_HELPER_CMD}} process cleanup-pid <pid-from-.prizmkit/state/browser-dev-server.pid>`
-3. Verify port is released: `{{RUNTIME_HELPER_CMD}} process cleanup-port $DEV_PORT`
+3. Verify port is released: `{{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}`
 {{END_IF_BROWSER_TOOL_PLAYWRIGHT}}
 {{IF_BROWSER_TOOL_OPENCLI}}
 **Using: opencli** (reuses Chrome logged-in sessions)
@@ -297,7 +297,7 @@ Run page-changing commands sequentially, then run `opencli browser state` to ref
 **Step 3 — Cleanup**:
 1. `opencli browser close`
 2. `{{RUNTIME_HELPER_CMD}} process cleanup-pid <pid-from-.prizmkit/state/browser-dev-server.pid>`
-3. `{{RUNTIME_HELPER_CMD}} process cleanup-port $DEV_PORT`
+3. `{{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}`
 {{END_IF_BROWSER_TOOL_OPENCLI}}
 {{IF_BROWSER_TOOL_AUTO}}
 **Tool Selection**: Choose the best browser tool at runtime.
@@ -328,7 +328,7 @@ Append results to `context-snapshot.md`:
    ```
    ## Browser Verification
    Tool: <playwright-cli or opencli>
-   URL: http://localhost:$DEV_PORT
+   URL: http://localhost:{{DEV_PORT}}
    Dev Server Command: <actual command used>
    Tool version: <version>
    Steps executed: [list of commands used]
@@ -366,7 +366,7 @@ This single commit includes: feature code + tests + .prizmkit/prizm-docs/ update
 ```bash
 git status --short
 ```
-Working tree MUST be clean after this step. If any feature-related files remain, stage them into the SAME commit via `git add <file> && git commit --amend --no-edit`, do NOT create a separate commit.
+Working tree MUST be clean after this step. If any feature-related files remain, stage them explicitly and amend the SAME commit with `git commit --amend --no-edit`; do NOT create a separate commit.
 
 **Exception**: `session-summary.md` in the artifact directory is a local cross-session artifact generated by `/prizmkit-committer` — it is NOT committed to git. Ignore it in the clean-tree check.
 

package/bundled/dev-pipeline/templates/bootstrap-tier2.md

@@ -52,15 +52,15 @@ You are running in **headless non-interactive mode** with a FINITE context windo
 
 1. **context-snapshot.md is your single source of truth** — After Phase 1 builds it, ALWAYS read context-snapshot.md instead of re-reading individual source files
 2. **Never re-read your own writes** — After you create/modify a file, do NOT read it back to verify. Trust your write was correct.
-3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use `grep -n '<pattern>' <file>` or `sed -n '<start>,<end>p' <file>` for targeted line lookups instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
+3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use platform file/search tools for targeted lookup; when shell-free helper lookup is needed, run `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<pattern>" --json`, then read only the smallest bounded range with the platform file-reading tool instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
 4. **Stay focused** — Do NOT explore code unrelated to this feature. No curiosity-driven reads.
 4. **One task at a time** — In Phase 4 (implement), complete and test one task before starting the next.
-5. **Minimize tool output** — Never load full command output into context. First capture to a temp file (`cmd 2>&1 | tee /tmp/out.txt | tail -20`), then scan the head/tail to identify relevant fields, and use targeted filtering (`grep`, `sed`, `awk`) to extract only the information needed for the current task. Only read the filtered result — never the raw full output.
+5. **Minimize tool output** — Never load full command output into context. First capture command output to a temp file using the command/tool output option or the host-appropriate redirection, then inspect only the final relevant lines and use targeted search helpers to extract the information needed for the current task. Only read the filtered result — never the raw full output.
 6. **No intermediate commits** — Do NOT run `git add`/`git commit` during Phase 1-5. All changes are committed once at the end in Phase 6 via `/prizmkit-committer`.
-7. **Capture test output once** — When running test suites, always use `<test-command> 2>&1 | tee /tmp/test-out.txt | tail -20`. Then grep `/tmp/test-out.txt` for details. Never re-run the suite just to apply a different filter.
+7. **Capture test output once** — When running test suites, capture output once to `/tmp/test-out.txt` using the command/tool output option or the host-appropriate redirection, surface only the final relevant lines, then search `/tmp/test-out.txt` for details. Never re-run the suite just to apply a different filter.
 9. **Package version verification (BLOCKING)** — Verify dependency versions via registry query before writing them. NEVER guess.
 10. **Build artifact hygiene** — After build/compile, ensure output is in `.gitignore`.
-11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Run `wc -l <file>` to get real line count, then `sed -n` for exact range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets — `grep -n '<anchor>'` first.
+11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Use the platform file-reading tool to refresh bounded line information, or locate an anchor with `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<anchor>" --json` before reading a small bounded range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets or prescribe POSIX-only lookup commands.
 
 ---
 
@@ -101,10 +101,10 @@ If any agent times out:
 
 **Check for previous failure log:**
 ```bash
-cat .prizmkit/specs/{{FEATURE_SLUG}}/failure-log.md 2>/dev/null || echo "NO_PREVIOUS_FAILURE"
+{{RUNTIME_HELPER_CMD}} artifact exists .prizmkit/specs/{{FEATURE_SLUG}}/failure-log.md --exists-marker PREVIOUS_FAILURE --missing-marker NO_PREVIOUS_FAILURE
 ```
-If failure-log.md exists:
-- Read ROOT_CAUSE and SUGGESTION — adjust your approach accordingly
+If the helper prints `PREVIOUS_FAILURE`:
+- Read `failure-log.md` ROOT_CAUSE and SUGGESTION — adjust your approach accordingly
 - Read DISCOVERED_TRAPS — if any are genuine, inject into .prizmkit/prizm-docs/ during Phase 6 retrospective
 - Do NOT delete failure-log.md until this session completes all phases and commits successfully
 
@@ -122,9 +122,9 @@ If MISSING — build it now:
 3. Scan the detected source directories for files related to this feature; read each one
 4. Write `.prizmkit/specs/{{FEATURE_SLUG}}/context-snapshot.md`:
    - **Section 1 — Feature Brief**: feature description + acceptance criteria (copy from above)
-   - **Section 2 — Project Structure**: run the following to get a visual directory tree, then paste output:
+   - **Section 2 — Project Structure**: run the helper to get a visual directory tree, then paste output:
      ```bash
-     find . -maxdepth 2 -type d -not -path '*/node_modules/*' -not -path '*/.git/*' -not -path '*/dist/*' -not -path '*/build/*' -not -path '*/__pycache__/*' -not -path '*/vendor/*' | sed -e 's;[^/]*/;|____;g;s;____|; |;g'
+     {{RUNTIME_HELPER_CMD}} artifact tree . --max-depth 2
      ```
    - **Section 3 — Key TRAPS & RULES**: relevant TRAPS/RULES from prizm-docs (not full copies)
    - **Section 4 — File Manifest**: For each file relevant to this feature, list: file path, why it's needed (modify/reference/test), key interface signatures (function names + params + return types). Do NOT include full file content — agents read files on-demand. Format:
@@ -264,7 +264,7 @@ After all critics return, read all 3 reports:
 
 Before starting, check plan.md Tasks:
 ```bash
-grep -c '^\- \[ \]' .prizmkit/specs/{{FEATURE_SLUG}}/plan.md 2>/dev/null || true
+{{RUNTIME_HELPER_CMD}} text count .prizmkit/specs/{{FEATURE_SLUG}}/plan.md --pattern "- [ ]"
 ```
 - If result is `0` (all tasks already `[x]`) → **SKIP to Review**.
 - If non-zero → implement directly below.
@@ -383,7 +383,7 @@ playwright-cli --help
 ```bash
 {{RUNTIME_HELPER_CMD}} platform skill playwright-cli --prefix
 ```
-If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy installed skill from `$HOME/.claude/skills/playwright-cli` to the path shown by the helper output for the current platform.
+If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy the installed skill from the Claude user skills directory to the path shown by the helper output for the current platform.
 
 0d. Read the installed playwright-cli skill (SKILL.md) for workflow guidance. Use its recommended patterns to construct your verification flow.
 
@@ -395,14 +395,14 @@ You know this project's tech stack. Detect and start the dev server yourself:
 2. Determine the dev server port from the pre-detected pipeline value (`{{DEV_PORT}}`) or project config. Do NOT hardcode or guess the port. Record it as `DEV_PORT` in your notes.
 3. Verify the port is available:
    ```bash
-   {{RUNTIME_HELPER_CMD}} port check $DEV_PORT
+   {{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}
    ```
 4. Start the dev server in background, capture PID:
    ```bash
    {{RUNTIME_HELPER_CMD}} process start --pid-file .prizmkit/state/browser-dev-server.pid -- <start-command tokens>
    ```
-5. Wait for server to be ready: run `{{RUNTIME_HELPER_CMD}} web wait-ready --url http://localhost:$DEV_PORT --timeout 30 --interval 2`
-6. Open the app in playwright-cli: `playwright-cli open http://localhost:$DEV_PORT`
+5. Wait for server to be ready: run `{{RUNTIME_HELPER_CMD}} web wait-ready --url http://localhost:{{DEV_PORT}} --timeout 30 --interval 2`
+6. Open the app in playwright-cli: `playwright-cli open http://localhost:{{DEV_PORT}}`
 7. If the page requires authentication, use playwright-cli to register a test user and log in first
 
 **Step 2 — Verification**:
@@ -416,7 +416,7 @@ Construct your verification workflow based on: (1) the playwright-cli skill docu
 
 1. Close the playwright-cli browser: `playwright-cli close`
 2. Kill the dev server process: `{{RUNTIME_HELPER_CMD}} process cleanup-pid <pid-from-.prizmkit/state/browser-dev-server.pid>`
-3. Verify port is released: `{{RUNTIME_HELPER_CMD}} process cleanup-port $DEV_PORT`
+3. Verify port is released: `{{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}`
 {{END_IF_BROWSER_TOOL_PLAYWRIGHT}}
 {{IF_BROWSER_TOOL_OPENCLI}}
 **Using: opencli** (reuses Chrome logged-in sessions)
@@ -454,7 +454,7 @@ Run page-changing commands sequentially, then run `opencli browser state` to ref
 **Step 3 — Cleanup**:
 1. `opencli browser close`
 2. `{{RUNTIME_HELPER_CMD}} process cleanup-pid <pid-from-.prizmkit/state/browser-dev-server.pid>`
-3. `{{RUNTIME_HELPER_CMD}} process cleanup-port $DEV_PORT`
+3. `{{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}`
 {{END_IF_BROWSER_TOOL_OPENCLI}}
 {{IF_BROWSER_TOOL_AUTO}}
 **Tool Selection**: Choose the best browser tool at runtime.
@@ -485,7 +485,7 @@ Append results to `context-snapshot.md`:
    ```
    ## Browser Verification
    Tool: <playwright-cli or opencli>
-   URL: http://localhost:$DEV_PORT
+   URL: http://localhost:{{DEV_PORT}}
    Dev Server Command: <actual command used>
    Tool version: <version>
    Steps executed: [list of commands used]
@@ -523,7 +523,7 @@ This single commit includes: feature code + tests + .prizmkit/prizm-docs/ update
 ```bash
 git status --short
 ```
-Working tree MUST be clean after this step. If any feature-related files remain, stage them into the SAME commit via `git add <file> && git commit --amend --no-edit`, do NOT create a separate commit.
+Working tree MUST be clean after this step. If any feature-related files remain, stage them explicitly and amend the SAME commit with `git commit --amend --no-edit`; do NOT create a separate commit.
 
 **Exception**: `session-summary.md` in the artifact directory is a local cross-session artifact generated by `/prizmkit-committer` — it is NOT committed to git. Ignore it in the clean-tree check.
 

package/bundled/dev-pipeline/templates/bootstrap-tier3.md

@@ -52,15 +52,15 @@ You are running in **headless non-interactive mode** with a FINITE context windo
 
 1. **context-snapshot.md is your single source of truth** — After Phase 1-2 builds it, ALWAYS read context-snapshot.md instead of re-reading individual source files
 2. **Never re-read your own writes** — After you create/modify a file, do NOT read it back to verify. Trust your write was correct.
-3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use `grep -n '<pattern>' <file>` or `sed -n '<start>,<end>p' <file>` for targeted line lookups instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
+3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use platform file/search tools for targeted lookup; when shell-free helper lookup is needed, run `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<pattern>" --json`, then read only the smallest bounded range with the platform file-reading tool instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
 4. **Stay focused** — Do NOT explore code unrelated to this feature. No curiosity-driven reads.
 4. **One task at a time** — In Phase 4 (implement), complete and test one task before starting the next.
-5. **Minimize tool output** — Never load full command output into context. First capture to a temp file (`cmd 2>&1 | tee /tmp/out.txt | tail -20`), then scan the head/tail to identify relevant fields, and use targeted filtering (`grep`, `sed`, `awk`) to extract only the information needed for the current task. Only read the filtered result — never the raw full output.
+5. **Minimize tool output** — Never load full command output into context. First capture command output to a temp file using the command/tool output option or the host-appropriate redirection, then inspect only the final relevant lines and use targeted search helpers to extract the information needed for the current task. Only read the filtered result — never the raw full output.
 6. **No intermediate commits** — Do NOT run `git add`/`git commit` during Phase 1-5. All changes are committed once at the end in Phase 6 via `/prizmkit-committer`.
 7. **Batch independent operations** — Issue multiple independent `Write`/`Read` calls in a single message turn when they have no dependencies. Combine multiple `mkdir -p` into one command. Never run `npm test` twice just to apply a different grep filter — capture output to `/tmp/test-out.txt` once and grep the file.
 9. **Package version verification (BLOCKING)** — Verify dependency versions via registry query before writing them. NEVER guess.
 10. **Build artifact hygiene** — After build/compile, ensure output is in `.gitignore`.
-11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Run `wc -l <file>` to get real line count, then `sed -n` for exact range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets — `grep -n '<anchor>'` first.
+11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Use the platform file-reading tool to refresh bounded line information, or locate an anchor with `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<anchor>" --json` before reading a small bounded range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets or prescribe POSIX-only lookup commands.
 
 ---
 
@@ -120,12 +120,12 @@ Check existing artifacts first:
 - `context-snapshot.md` exists → use it directly, skip Phase 1
 - Some missing → generate only missing files
 
-Before planning, check whether feature code already exists in the project (search in source directories identified from `root.prizm` or the project tree scan):
+Before planning, check whether feature code already exists in the project (use the helper to search source directories identified from `root.prizm` or the project tree scan):
 ```bash
-grep -r "{{FEATURE_SLUG}}" . --include="*.js" --include="*.ts" --include="*.py" --include="*.go" --include="*.java" --include="*.rb" --include="*.rs" -l --exclude-dir=node_modules --exclude-dir=.git --exclude-dir=dist --exclude-dir=build --exclude-dir=vendor --exclude-dir=.prizmkit 2>/dev/null | head -20
+{{RUNTIME_HELPER_CMD}} text search . --pattern "{{FEATURE_SLUG}}" --json
 ```
 
-Record result as `EXISTING_CODE` (list of files, or empty).
+Record result as `EXISTING_CODE` (helper `match` entries for source files, or empty; ignore `.prizmkit/` artifact matches).
 
 If `EXISTING_CODE` is non-empty: your spec/plan/tasks must reflect this existing implementation — document what exists, identify gaps, do NOT re-implement what is already done.
 
@@ -259,7 +259,7 @@ After all critics return, read all 3 reports:
 
 Before starting, check plan.md Tasks:
 ```bash
-grep -c '^\- \[ \]' .prizmkit/specs/{{FEATURE_SLUG}}/plan.md 2>/dev/null || true
+{{RUNTIME_HELPER_CMD}} text count .prizmkit/specs/{{FEATURE_SLUG}}/plan.md --pattern "- [ ]"
 ```
 - If result is `0` (all tasks already `[x]`) → **SKIP to Review**.
 - If non-zero → implement directly below.
@@ -378,7 +378,7 @@ playwright-cli --help
 ```bash
 {{RUNTIME_HELPER_CMD}} platform skill playwright-cli --prefix
 ```
-If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy installed skill from `$HOME/.claude/skills/playwright-cli` to the path shown by the helper output for the current platform.
+If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy the installed skill from the Claude user skills directory to the path shown by the helper output for the current platform.
 
 0d. Read the installed playwright-cli skill (SKILL.md) for workflow guidance. Use its recommended patterns to construct your verification flow.
 
@@ -390,14 +390,14 @@ You know this project's tech stack. Detect and start the dev server yourself:
 2. Determine the dev server port from the pre-detected pipeline value (`{{DEV_PORT}}`) or project config. Do NOT hardcode or guess the port. Record it as `DEV_PORT` in your notes.
 3. Verify the port is available:
    ```bash
-   {{RUNTIME_HELPER_CMD}} port check $DEV_PORT
+   {{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}
    ```
 4. Start the dev server in background, capture PID:
    ```bash
    {{RUNTIME_HELPER_CMD}} process start --pid-file .prizmkit/state/browser-dev-server.pid -- <start-command tokens>
    ```
-5. Wait for server to be ready: run `{{RUNTIME_HELPER_CMD}} web wait-ready --url http://localhost:$DEV_PORT --timeout 30 --interval 2`
-6. Open the app in playwright-cli: `playwright-cli open http://localhost:$DEV_PORT`
+5. Wait for server to be ready: run `{{RUNTIME_HELPER_CMD}} web wait-ready --url http://localhost:{{DEV_PORT}} --timeout 30 --interval 2`
+6. Open the app in playwright-cli: `playwright-cli open http://localhost:{{DEV_PORT}}`
 7. If the page requires authentication, use playwright-cli to register a test user and log in first
 
 **Step 2 — Verification**:
@@ -411,7 +411,7 @@ Construct your verification workflow based on: (1) the playwright-cli skill docu
 
 1. Close the playwright-cli browser: `playwright-cli close`
 2. Kill the dev server process: `{{RUNTIME_HELPER_CMD}} process cleanup-pid <pid-from-.prizmkit/state/browser-dev-server.pid>`
-3. Verify port is released: `{{RUNTIME_HELPER_CMD}} process cleanup-port $DEV_PORT`
+3. Verify port is released: `{{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}`
 {{END_IF_BROWSER_TOOL_PLAYWRIGHT}}
 {{IF_BROWSER_TOOL_OPENCLI}}
 **Using: opencli** (reuses Chrome logged-in sessions)
@@ -449,7 +449,7 @@ Run page-changing commands sequentially, then run `opencli browser state` to ref
 **Step 3 — Cleanup**:
 1. `opencli browser close`
 2. `{{RUNTIME_HELPER_CMD}} process cleanup-pid <pid-from-.prizmkit/state/browser-dev-server.pid>`
-3. `{{RUNTIME_HELPER_CMD}} process cleanup-port $DEV_PORT`
+3. `{{RUNTIME_HELPER_CMD}} port check {{DEV_PORT}}`
 {{END_IF_BROWSER_TOOL_OPENCLI}}
 {{IF_BROWSER_TOOL_AUTO}}
 **Tool Selection**: Choose the best browser tool at runtime.
@@ -480,7 +480,7 @@ Append results to `context-snapshot.md`:
    ```
    ## Browser Verification
    Tool: <playwright-cli or opencli>
-   URL: http://localhost:$DEV_PORT
+   URL: http://localhost:{{DEV_PORT}}
    Dev Server Command: <actual command used>
    Tool version: <version>
    Steps executed: [list of commands used]
@@ -502,9 +502,9 @@ If verification fails, log the failure details but continue to commit. Failures
 - Complex bugs (multi-module, cascading): Use `/prizmkit-plan` with `artifact_dir=.prizmkit/bugfix/<BUG_ID>/`.
 - Commit prefix: `fix(<scope>):` (not `feat:`).
 
-**6a.** Check if feature already committed:
+**6a.** Check if feature already committed by inspecting recent git history for `{{FEATURE_ID}}`:
 ```bash
-git log --oneline | grep "{{FEATURE_ID}}" | head -3
+git log --oneline
 ```
 - If a commit for `{{FEATURE_ID}}` already exists → **skip 6d** (do NOT run /prizmkit-committer, do NOT run git reset, do NOT stage or unstage anything). Proceed directly to 6e Final verification.
 - If no existing commit → proceed normally with 6b–6d.
@@ -533,7 +533,7 @@ This single commit includes: feature code + tests + .prizmkit/prizm-docs/ update
 ```bash
 git status --short
 ```
-Working tree MUST be clean after this step. If any feature-related files remain, stage them into the SAME commit via `git add <file> && git commit --amend --no-edit`, do NOT create a separate commit.
+Working tree MUST be clean after this step. If any feature-related files remain, stage them explicitly and amend the SAME commit with `git commit --amend --no-edit`; do NOT create a separate commit.
 
 **Exception**: `session-summary.md` in the artifact directory is a local cross-session artifact generated by `/prizmkit-committer` — it is NOT committed to git. Ignore it in the clean-tree check.
 

package/bundled/dev-pipeline/templates/sections/checkpoint-system.md

@@ -8,7 +8,7 @@ A checkpoint file tracks your progress through this workflow:
 
 Use this helper to update checkpoint status. Do not hand-edit `workflow-checkpoint.json` directly.
 
-```bash
+```text
 {{CHECKPOINT_PYTHON_CMD}} \
   --checkpoint-path {{CHECKPOINT_PATH}} \
   --step <step-id-or-skill-name> \

package/bundled/dev-pipeline/templates/sections/context-budget-rules.md

@@ -6,18 +6,18 @@ You are running in **headless non-interactive mode** with a FINITE context windo
 
 1. **context-snapshot.md is your single source of truth** — After it is built, ALWAYS read context-snapshot.md instead of re-reading individual source files
 2. **Never re-read your own writes** — After you create/modify a file, do NOT read it back to verify. Trust your write was correct.
-3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use `grep -n '<pattern>' <file>` or `sed -n '<start>,<end>p' <file>` for targeted line lookups instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
+3. **Never re-read unmodified files** — Once you have read a file into context, do NOT read it again unless it was modified by an Edit/Write operation after your last read. Use platform file/search tools for targeted lookup; when shell-free helper lookup is needed, run `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<pattern>" --json`, then read only the smallest bounded range with the platform file-reading tool instead of full-file Read. Before issuing a Read, ask yourself: "Have I read this file before, and has it changed since?" If the answer is "no, it hasn't changed" — do NOT read it. This applies to all files including those summarized in context-snapshot.md Section 4. Subagents (Critic, Reviewer) must receive context-snapshot.md and rely on it rather than independently re-reading the same files the orchestrator already read.
 4. **Stay focused** — Do NOT explore code unrelated to this feature. No curiosity-driven reads.
 5. **One task at a time** — Complete and test one task before starting the next.
-6. **Minimize tool output** — Never load full command output into context. First capture to a temp file (`cmd 2>&1 | tee /tmp/out.txt | tail -20`), then scan the head/tail to identify relevant fields, and use targeted filtering (`grep`, `sed`, `awk`) to extract only the information needed for the current task. Only read the filtered result — never the raw full output.
+6. **Minimize tool output** — Never load full command output into context. First capture command output to a temp file using the command/tool output option or the host-appropriate redirection, then inspect only the final relevant lines and use targeted search helpers to extract the information needed for the current task. Only read the filtered result — never the raw full output.
 7. **No intermediate commits** — Do NOT run `git add`/`git commit` during implementation phases. All changes are committed once at the end via `/prizmkit-committer`.
-8. **Capture test output once** — When running test suites during the post-review gate, capture with `<test-command> 2>&1 | tee /tmp/test-out.txt | tail -20`. Then grep `/tmp/test-out.txt` for details. Never re-run the suite just to apply a different filter.
+8. **Capture test output once** — When running test suites during the post-review gate, capture output once to `/tmp/test-out.txt` using the command/tool output option or the host-appropriate redirection, surface only the final relevant lines, then search `/tmp/test-out.txt` for details. Never re-run the suite just to apply a different filter.
 9. **Package version verification (HARD CONSTRAINT — BLOCKING)** — Before writing ANY dependency version in `package.json`, `requirements.txt`, `Cargo.toml`, `go.mod`, `Gemfile`, `pyproject.toml`, or any other dependency manifest:
    - You MUST verify the real version exists by querying the package registry first:
-     - npm/Node.js: `npm view <package> dist-tags.latest 2>/dev/null`
-     - Python/pip: `pip index versions <package> 2>/dev/null | head -1`
-     - Go: `go list -m -versions <module>@latest 2>/dev/null`
-     - Rust: `cargo search <crate> --limit 1 2>/dev/null`
+     - npm/Node.js: `npm view <package> dist-tags.latest`
+     - Python/pip: `pip index versions <package>` and use the first result line
+     - Go: `go list -m -versions <module>@latest`
+     - Rust: `cargo search <crate> --limit 1`
    - **NEVER guess or hallucinate version numbers**. If you cannot verify a version, use `"latest"` or `"*"` as a placeholder, or omit the version constraint entirely and let the package manager resolve it
    - If the registry query fails (network issue, package not found), you MUST either:
      (a) Use a known-safe version you have high confidence in, OR
@@ -26,4 +26,4 @@ You are running in **headless non-interactive mode** with a FINITE context windo
    - **This is a BLOCKING gate**: do NOT run `npm install` / `pip install` / `cargo build` / `go mod tidy` until ALL versions in the manifest have been verified or use open constraints
    - Batch version lookups: query multiple packages in parallel to save time (e.g., run multiple `npm view` commands concurrently)
 10. **Build artifact hygiene** — After any build/compile command (`go build`, `npm run build`, `tsc`, etc.), ensure the output binary, build directory, generated bundle, or cache directory is ignored by git before the commit phase. Never commit compiled binaries, build output, or generated artifacts.
-11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Run `wc -l <file>` to get real line count, then `sed -n '<start>,<end>p' <file>` for exact range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets — `grep -n '<anchor>'` first.
+11. **Read offset safety (BLOCKING)** — If Read returns "shorter than offset"/"File is empty"/"Wasted call" — do NOT retry same offset. Use the platform file-reading tool to refresh bounded line information, or locate an anchor with `{{RUNTIME_HELPER_CMD}} text search <file> --pattern "<anchor>" --json` before reading a small bounded range. 2 consecutive offset errors → STOP Reading that file. Never guess offsets or prescribe POSIX-only lookup commands.

package/bundled/dev-pipeline/templates/sections/phase-browser-verification-auto.md

@@ -34,7 +34,7 @@ If tool is `skipped`, do NOT start a dev server and do NOT run browser commands.
 ```bash
 {{RUNTIME_HELPER_CMD}} platform skill playwright-cli --prefix
 ```
-If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy installed skill from `$HOME/.claude/skills/playwright-cli` to the path shown by the helper output for the current platform.
+If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy the installed skill from the Claude user skills directory to the path shown by the helper output for the current platform.
 
 0e. Read the installed playwright-cli skill (SKILL.md) for workflow guidance. Learn usage: `playwright-cli --help`.
 
@@ -62,7 +62,8 @@ Construct your verification workflow based on: (1) the playwright-cli skill docu
 
 **Step 1 — Start Dev Server + Open (opencli)**:
 1. Identify and start the dev server (see shared helper commands below)
-2. Open and inspect: `opencli browser open http://localhost:<DEV_PORT> && opencli browser state`
+2. Open: `opencli browser open http://localhost:<DEV_PORT>`
+3. Inspect: `opencli browser state`
 3. If auth needed, opencli reuses Chrome cookies — SSO/OAuth may already be active
 
 **Step 2 — Verification (opencli)**:
@@ -71,7 +72,7 @@ Use `opencli browser state` to discover elements with `[N]` indices, then verify
 
 Key commands: `state`, `click <N>`, `type <N> "text"`, `get text <N>`, `get value <N>`, `wait text "..."`, `wait selector "..."`, `scroll down`, `eval "(function(){ ... })()"`.
 
-**Chain commands with `&&`**: `opencli browser type 3 "hello" && opencli browser click 7`
+Run page-changing commands sequentially; for example, run `opencli browser type 3 "hello"`, then run `opencli browser click 7`.
 
 Always run `state` after page-changing actions to get fresh indices.
 

package/bundled/dev-pipeline/templates/sections/phase-browser-verification-opencli.md

@@ -51,7 +51,8 @@ You know this project's tech stack. Detect and start the dev server yourself:
    ```
 6. Open the app in opencli and inspect page state:
    ```bash
-   opencli browser open http://localhost:<DEV_PORT> && opencli browser state
+   opencli browser open http://localhost:<DEV_PORT>
+   opencli browser state
    ```
 7. If the page requires authentication, use opencli browser to interact with login forms (opencli reuses Chrome cookies, so SSO/OAuth may already be active)
 
@@ -71,11 +72,14 @@ Key opencli browser commands for verification:
 - `opencli browser scroll down` — scroll page
 - `opencli browser eval "(function(){ ... })()"` — read-only JS evaluation for data extraction
 
-**Chain commands aggressively with `&&`** to minimize tool calls:
+**Run commands sequentially and refresh state after page-changing actions**:
 ```bash
-opencli browser open http://localhost:<DEV_PORT> && opencli browser state
-opencli browser type 3 "hello" && opencli browser type 4 "world" && opencli browser click 7
-opencli browser click 12 && opencli browser wait time 1 && opencli browser state
+opencli browser open http://localhost:<DEV_PORT>
+opencli browser state
+opencli browser type 3 "hello"
+opencli browser type 4 "world"
+opencli browser click 7
+opencli browser state
 ```
 
 **IMPORTANT**: Always run `opencli browser state` after page-changing actions (open, click on links, scroll) to get fresh element indices. Never guess indices.

package/bundled/dev-pipeline/templates/sections/phase-browser-verification.md

@@ -34,7 +34,7 @@ Use this output to determine the correct commands for your verification steps. D
 ```bash
 {{RUNTIME_HELPER_CMD}} platform skill playwright-cli --prefix
 ```
-If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy installed skill from `$HOME/.claude/skills/playwright-cli` to the path shown by the helper output for the current platform.
+If `SKILL_MISSING`: run `playwright-cli install --skills`. If current platform is NOT claude, copy the installed skill from the Claude user skills directory to the path shown by the helper output for the current platform.
 
 0d. Read the installed playwright-cli skill for workflow guidance:
 After skill installation, read the skill's SKILL.md to understand recommended workflows and patterns. Use these patterns to construct your verification flow — do NOT invent your own patterns if the skill provides them.

package/bundled/dev-pipeline/templates/sections/phase-commit-full.md

@@ -2,7 +2,7 @@
 
 **a.** Check if feature already committed:
 ```bash
-git log --oneline | grep "{{FEATURE_ID}}" | head -3
+git log --oneline
 ```
 - If a commit for `{{FEATURE_ID}}` already exists → **skip d** (do NOT run /prizmkit-committer, do NOT run git reset, do NOT stage or unstage anything). Proceed directly to e Final verification.
 - If no existing commit → proceed normally with b–d.
@@ -30,7 +30,7 @@ This single commit includes: feature code + tests + .prizmkit/prizm-docs/ update
 ```bash
 git status --short
 ```
-Working tree MUST be clean after this step. If any feature-related files remain, stage them into the SAME commit via `git add <file> && git commit --amend --no-edit`, do NOT create a separate commit.
+Working tree MUST be clean after this step. If any feature-related files remain, stage them explicitly and amend the SAME commit with `git commit --amend --no-edit`; do NOT create a separate commit.
 
 **f.** Write completion summary for downstream dependency context:
 

package/bundled/dev-pipeline/templates/sections/phase-commit.md

@@ -23,7 +23,7 @@ This single commit includes: feature code + tests + .prizmkit/prizm-docs/ update
 ```bash
 git status --short
 ```
-Working tree MUST be clean after this step. If any feature-related files remain, stage them into the SAME commit via `git add <file> && git commit --amend --no-edit`, do NOT create a separate commit.
+Working tree MUST be clean after this step. If any feature-related files remain, stage them explicitly and amend the SAME commit with `git commit --amend --no-edit`; do NOT create a separate commit.
 
 **e.** Write completion summary for downstream dependency context:
 

package/bundled/dev-pipeline/templates/sections/phase-implement-agent.md

@@ -4,7 +4,7 @@
 
 Before starting, check plan.md Tasks:
 ```bash
-grep -c '^\- \[ \]' .prizmkit/specs/{{FEATURE_SLUG}}/plan.md 2>/dev/null || true
+{{RUNTIME_HELPER_CMD}} text count .prizmkit/specs/{{FEATURE_SLUG}}/plan.md --pattern "- [ ]"
 ```
 - If result is `0` (all tasks already `[x]`) → **SKIP to Review**.
 - If non-zero → implement directly below.

package/bundled/dev-pipeline/templates/sections/phase-implement-full.md

@@ -4,7 +4,7 @@
 
 Before starting, check plan.md Tasks:
 ```bash
-grep -c '^\- \[ \]' .prizmkit/specs/{{FEATURE_SLUG}}/plan.md 2>/dev/null || true
+{{RUNTIME_HELPER_CMD}} text count .prizmkit/specs/{{FEATURE_SLUG}}/plan.md --pattern "- [ ]"
 ```
 - If result is `0` (all tasks already `[x]`) → **SKIP to Review**.
 - If non-zero → implement directly below.