spec-runner 2.3.0__tar.gz → 2.4.0__tar.gz

{spec_runner-2.3.0/src/spec_runner.egg-info → spec_runner-2.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: spec-runner
-Version: 2.3.0
+Version: 2.4.0
 Summary: Task automation from markdown specs via Claude CLI
 Author: Andrei
 License-Expression: MIT
@@ -134,7 +134,8 @@ Tasks are defined in `spec/tasks.md`:
 # Execution
 spec-runner run                            # Execute next ready task
 spec-runner run --task=TASK-001            # Execute specific task
-spec-runner run --all                      # Execute all ready tasks
+spec-runner run --all                      # Execute all ready tasks (resets failed→pending by default)
+spec-runner run --all --no-reset-failed    # Keep failed tasks sticky (skip the default reset)
 spec-runner run --all --hitl-review        # Interactive HITL approval gate
 spec-runner run --force                    # Skip lock check (stale lock)
 spec-runner run --tui                      # Execute with live TUI dashboard
@@ -177,6 +178,14 @@ spec-runner report --json                  # JSON matrix output
 # Planning
 spec-runner plan "description"             # Interactive task planning
 spec-runner plan --full "description"      # Generate full spec (requirements + design + tasks)
+spec-runner plan --full --from-file spec.md  # Read the description from a file instead of an arg
+
+# Diagnostics
+spec-runner doctor                              # Probe the configured CLI/model (real mini-task)
+spec-runner doctor --cli=codex --model=gpt-5.4  # Probe an ad-hoc CLI+model
+spec-runner doctor --with-review                # Also probe the review stage
+spec-runner doctor --json --yes                 # Machine-readable, no confirmation (CI)
+spec-runner doctor --strict                     # Exit non-zero on DEGRADED too
 
 # Integration
 spec-runner mcp                            # Launch MCP server (stdio)
@@ -368,13 +377,34 @@ paths:
 | CLI | Auto-detected | Example template |
 |-----|--------------|------------------|
 | Claude | Yes | `{cmd} -p {prompt} --model {model}` |
-| Codex | Yes | `{cmd} -p {prompt} --model {model}` |
+| Codex | Yes | `{cmd} exec -m {model} {prompt}` (codex's `-p` is `--profile`, not the prompt) |
 | OpenCode ([sst/opencode](https://opencode.ai)) | Yes | `{cmd} run --model {model} {prompt}` |
 | Pi Agent ([pi.dev](https://pi.dev)) | Yes (basename match) | `{cmd} -p --model {model} {prompt}` |
 | Ollama | Yes | `{cmd} run {model} {prompt}` |
 | llama-cli | Yes | `{cmd} -m {model} -p {prompt} --no-display-prompt` |
 | Custom | Use template | `{cmd} --prompt {prompt}` |
 
+> **Full pi-driven loop:** `pi` can run the entire dev → review → test cycle (with native
+> skills, per-stage tool control and a read-only review gate) using only config and a small
+> script — no core code. See [docs/pi-workflow.md](docs/pi-workflow.md) and the runnable
+> [examples/pi-loop/](examples/pi-loop/).
+
+### Checking CLI/model compatibility
+
+`spec-runner doctor` runs a real one-task probe through the actual execution
+path and reports, per capability, whether your CLI/model works:
+
+- **invocation** — the command runs and authenticates
+- **completion_marker** — the model prints `TASK_COMPLETE` (not all models do)
+- **task_action** — the model actually performs the work
+- **cost_tracking** — token/cost parsing works (needed for `costs`/`--budget`)
+- **error_classification** — failures are classified (diagnostic)
+- **review** *(with `--with-review`)* — the reviewer prints `REVIEW_PASSED`/`FAILED`
+
+Verdict: **READY** / **DEGRADED** (works, but something like cost tracking is
+unavailable) / **BROKEN**. It makes real, billable model calls (capped by
+`--budget`, default $0.50) and asks for confirmation unless `--yes`.
+
 ## Project Structure
 
 ```
@@ -391,6 +421,8 @@ project/
 │       ├── cli_info.py          # Status, costs, logs, validate, verify, report, TUI, MCP
 │       ├── cli_plan.py          # Interactive planning command
 │       ├── execution.py         # Task execution + retry logic
+│       ├── errors.py            # CLI stderr → human-readable failure reasons
+│       ├── stages.py            # Per-task sub-stage tracking (StageReporter)
 │       ├── config.py            # ExecutorConfig + YAML loading
 │       ├── state.py             # SQLite state persistence + degraded-mode fallback
 │       ├── prompt.py            # Prompt building + templates
@@ -407,7 +439,8 @@ project/
 │       ├── report.py            # Traceability matrix generation
 │       ├── validate.py          # Config + task validation
 │       ├── plugins.py           # Plugin discovery + hooks
-│       ├── logging.py           # Structured logging (structlog)
+│       ├── logging.py           # Structured logging (structlog back-compat shim)
+│       ├── obs.py               # OTel JSONL observability emitter (shared contract)
 │       ├── events.py            # EventBus for streaming to TUI
 │       ├── notifications.py     # Telegram + webhook notifications
 │       ├── tui.py               # Textual TUI dashboard

{spec_runner-2.3.0 → spec_runner-2.4.0}/README.md

@@ -99,7 +99,8 @@ Tasks are defined in `spec/tasks.md`:
 # Execution
 spec-runner run                            # Execute next ready task
 spec-runner run --task=TASK-001            # Execute specific task
-spec-runner run --all                      # Execute all ready tasks
+spec-runner run --all                      # Execute all ready tasks (resets failed→pending by default)
+spec-runner run --all --no-reset-failed    # Keep failed tasks sticky (skip the default reset)
 spec-runner run --all --hitl-review        # Interactive HITL approval gate
 spec-runner run --force                    # Skip lock check (stale lock)
 spec-runner run --tui                      # Execute with live TUI dashboard
@@ -142,6 +143,14 @@ spec-runner report --json                  # JSON matrix output
 # Planning
 spec-runner plan "description"             # Interactive task planning
 spec-runner plan --full "description"      # Generate full spec (requirements + design + tasks)
+spec-runner plan --full --from-file spec.md  # Read the description from a file instead of an arg
+
+# Diagnostics
+spec-runner doctor                              # Probe the configured CLI/model (real mini-task)
+spec-runner doctor --cli=codex --model=gpt-5.4  # Probe an ad-hoc CLI+model
+spec-runner doctor --with-review                # Also probe the review stage
+spec-runner doctor --json --yes                 # Machine-readable, no confirmation (CI)
+spec-runner doctor --strict                     # Exit non-zero on DEGRADED too
 
 # Integration
 spec-runner mcp                            # Launch MCP server (stdio)
@@ -333,13 +342,34 @@ paths:
 | CLI | Auto-detected | Example template |
 |-----|--------------|------------------|
 | Claude | Yes | `{cmd} -p {prompt} --model {model}` |
-| Codex | Yes | `{cmd} -p {prompt} --model {model}` |
+| Codex | Yes | `{cmd} exec -m {model} {prompt}` (codex's `-p` is `--profile`, not the prompt) |
 | OpenCode ([sst/opencode](https://opencode.ai)) | Yes | `{cmd} run --model {model} {prompt}` |
 | Pi Agent ([pi.dev](https://pi.dev)) | Yes (basename match) | `{cmd} -p --model {model} {prompt}` |
 | Ollama | Yes | `{cmd} run {model} {prompt}` |
 | llama-cli | Yes | `{cmd} -m {model} -p {prompt} --no-display-prompt` |
 | Custom | Use template | `{cmd} --prompt {prompt}` |
 
+> **Full pi-driven loop:** `pi` can run the entire dev → review → test cycle (with native
+> skills, per-stage tool control and a read-only review gate) using only config and a small
+> script — no core code. See [docs/pi-workflow.md](docs/pi-workflow.md) and the runnable
+> [examples/pi-loop/](examples/pi-loop/).
+
+### Checking CLI/model compatibility
+
+`spec-runner doctor` runs a real one-task probe through the actual execution
+path and reports, per capability, whether your CLI/model works:
+
+- **invocation** — the command runs and authenticates
+- **completion_marker** — the model prints `TASK_COMPLETE` (not all models do)
+- **task_action** — the model actually performs the work
+- **cost_tracking** — token/cost parsing works (needed for `costs`/`--budget`)
+- **error_classification** — failures are classified (diagnostic)
+- **review** *(with `--with-review`)* — the reviewer prints `REVIEW_PASSED`/`FAILED`
+
+Verdict: **READY** / **DEGRADED** (works, but something like cost tracking is
+unavailable) / **BROKEN**. It makes real, billable model calls (capped by
+`--budget`, default $0.50) and asks for confirmation unless `--yes`.
+
 ## Project Structure
 
 ```
@@ -356,6 +386,8 @@ project/
 │       ├── cli_info.py          # Status, costs, logs, validate, verify, report, TUI, MCP
 │       ├── cli_plan.py          # Interactive planning command
 │       ├── execution.py         # Task execution + retry logic
+│       ├── errors.py            # CLI stderr → human-readable failure reasons
+│       ├── stages.py            # Per-task sub-stage tracking (StageReporter)
 │       ├── config.py            # ExecutorConfig + YAML loading
 │       ├── state.py             # SQLite state persistence + degraded-mode fallback
 │       ├── prompt.py            # Prompt building + templates
@@ -372,7 +404,8 @@ project/
 │       ├── report.py            # Traceability matrix generation
 │       ├── validate.py          # Config + task validation
 │       ├── plugins.py           # Plugin discovery + hooks
-│       ├── logging.py           # Structured logging (structlog)
+│       ├── logging.py           # Structured logging (structlog back-compat shim)
+│       ├── obs.py               # OTel JSONL observability emitter (shared contract)
 │       ├── events.py            # EventBus for streaming to TUI
 │       ├── notifications.py     # Telegram + webhook notifications
 │       ├── tui.py               # Textual TUI dashboard

{spec_runner-2.3.0 → spec_runner-2.4.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "spec-runner"
-version = "2.3.0"
+version = "2.4.0"
 description = "Task automation from markdown specs via Claude CLI"
 readme = "README.md"
 requires-python = ">=3.10"

{spec_runner-2.3.0 → spec_runner-2.4.0}/src/spec_runner/cli.py

@@ -202,9 +202,7 @@ def _run_tasks(args, config: ExecutorConfig):
             tasks = parse_tasks(config.tasks_file)
 
         # v2.3.0: reset failed-task state on `run --all` unless opted out.
-        reset_enabled = getattr(args, "all", False) and not getattr(
-            args, "no_reset_failed", False
-        )
+        reset_enabled = getattr(args, "all", False) and not getattr(args, "no_reset_failed", False)
         previously_failed: set[str] = set()  # used by T17 second-pass detection
         if reset_enabled:
             previously_failed = state.reset_failed_to_pending()
@@ -416,8 +414,7 @@ def _run_tasks(args, config: ExecutorConfig):
                     else:
                         stop_reason = "max_consecutive_failures"
                         stop_detail = (
-                            f"{state.consecutive_failures}/"
-                            f"{config.max_consecutive_failures}"
+                            f"{state.consecutive_failures}/{config.max_consecutive_failures}"
                         )
                     logger.warning("Stopping: too many consecutive failures")
                     break
@@ -460,8 +457,7 @@ def _run_tasks(args, config: ExecutorConfig):
                     else:
                         stop_reason = "max_consecutive_failures"
                         stop_detail = (
-                            f"{state.consecutive_failures}/"
-                            f"{config.max_consecutive_failures}"
+                            f"{state.consecutive_failures}/{config.max_consecutive_failures}"
                         )
                     logger.warning("Stopping: too many consecutive failures")
                     break
@@ -663,6 +659,25 @@ def cmd_watch(args: argparse.Namespace, config: ExecutorConfig) -> None:
         time.sleep(1)
 
 
+def cmd_doctor(args: argparse.Namespace, config: ExecutorConfig) -> None:
+    """Run the CLI/model compatibility probe and exit with its status code."""
+    from .doctor import run_doctor
+
+    code = run_doctor(
+        config,
+        cli=args.cli,
+        model=args.model,
+        with_review=args.with_review,
+        budget=args.budget,
+        timeout_min=getattr(args, "timeout", None),
+        assume_yes=args.yes,
+        strict=args.strict,
+        as_json=args.json,
+        keep=args.keep,
+    )
+    raise SystemExit(code)
+
+
 # === Main ===
 
 
@@ -853,7 +868,14 @@ def _build_parser() -> argparse.ArgumentParser:
 
     # plan
     plan_parser = subparsers.add_parser("plan", parents=[common], help="Interactive task planning")
-    plan_parser.add_argument("description", help="Feature description")
+    plan_parser.add_argument(
+        "description", nargs="?", default=None, help="Feature description (or use --from-file)"
+    )
+    plan_parser.add_argument(
+        "--from-file",
+        metavar="PATH",
+        help="Read the feature description from a file instead of the positional argument",
+    )
     plan_parser.add_argument(
         "--full",
         action="store_true",
@@ -943,6 +965,28 @@ def _build_parser() -> argparse.ArgumentParser:
     # mcp
     subparsers.add_parser("mcp", parents=[common], help="Launch read-only MCP server")
 
+    # doctor
+    doctor_parser = subparsers.add_parser(
+        "doctor", parents=[common], help="Probe CLI/model compatibility (real mini-task)"
+    )
+    doctor_parser.add_argument("--cli", help="Override the CLI command (claude/codex/pi/...)")
+    doctor_parser.add_argument("--model", help="Override the model (executor + review)")
+    doctor_parser.add_argument(
+        "--with-review",
+        action="store_true",
+        help="Also probe the review stage (2nd model call)",
+    )
+    doctor_parser.add_argument(
+        "--yes", "-y", action="store_true", help="Skip the cost-gate confirmation"
+    )
+    doctor_parser.add_argument(
+        "--strict", action="store_true", help="Exit non-zero on DEGRADED too"
+    )
+    doctor_parser.add_argument("--json", action="store_true", help="Machine-readable output")
+    doctor_parser.add_argument("--keep", action="store_true", help="Keep the scratch workspace")
+    # --budget is inherited from common (default None); override default to 0.50 for doctor
+    doctor_parser.set_defaults(budget=0.5)
+
     # task (unified: replaces spec-task binary)
     task_parser = subparsers.add_parser(
         "task", help="Task management (list, show, start, done, graph, sync)"
@@ -1038,6 +1082,7 @@ def main():
         "tui": cmd_tui,
         "watch": cmd_watch,
         "mcp": cmd_mcp,
+        "doctor": cmd_doctor,
     }
 
     # Handle unified task subcommand

{spec_runner-2.3.0 → spec_runner-2.4.0}/src/spec_runner/cli_plan.py

@@ -4,6 +4,7 @@ import re
 import subprocess
 import sys
 from datetime import datetime
+from pathlib import Path
 
 from .config import ExecutorConfig
 from .logging import get_logger
@@ -23,6 +24,32 @@ from .task import (
 logger = get_logger("cli")
 
 
+def resolve_plan_description(description: str | None, from_file: str | None) -> str:
+    """Resolve the plan description from --from-file (preferred) or the positional
+    argument. Exits with an error if neither is usable.
+
+    Args:
+        description: the positional description (may be None when --from-file is used).
+        from_file: path to a file whose contents are the description.
+    """
+    if from_file:
+        path = Path(from_file)
+        if not path.is_file():
+            raise SystemExit(f"plan --from-file: not a readable file: {from_file}")
+        try:
+            text = path.read_text(encoding="utf-8").strip()
+        except UnicodeDecodeError as e:
+            raise SystemExit(f"plan --from-file: not valid UTF-8 text: {from_file}") from e
+        except OSError as e:
+            raise SystemExit(f"plan --from-file: cannot read {from_file}: {e}") from e
+        if not text:
+            raise SystemExit(f"plan --from-file: file is empty: {from_file}")
+        return text
+    if description and description.strip():
+        return description
+    raise SystemExit("plan: provide a description argument or --from-file PATH")
+
+
 def cmd_plan(args, config: ExecutorConfig):
     """Interactive task planning via Claude.
 
@@ -30,7 +57,7 @@ def cmd_plan(args, config: ExecutorConfig):
     requirements, design, and tasks files from a description.
     """
 
-    description = args.description
+    description = resolve_plan_description(args.description, getattr(args, "from_file", None))
 
     if getattr(args, "full", False):
         from .prompt import build_generation_prompt, parse_spec_marker

{spec_runner-2.3.0 → spec_runner-2.4.0}/src/spec_runner/config.py

@@ -153,6 +153,7 @@ class ExecutorConfig:
     create_git_branch: bool = True  # Create branch on start
     auto_commit: bool = True  # Auto-commit on success
     main_branch: str = ""  # Main branch name (empty = auto-detect: main/master)
+    sync_deps: bool = True  # Run `uv sync` in pre_start_hook (doctor disables this)
 
     # Code review
     run_review: bool = True  # Run code review after task completion
@@ -377,6 +378,7 @@ def load_config_from_yaml(config_path: Path | None = None) -> dict:
             "claude_model": executor_config.get("claude_model"),
             "skip_permissions": executor_config.get("skip_permissions"),
             "create_git_branch": pre_start.get("create_git_branch"),
+            "sync_deps": pre_start.get("sync_deps"),
             "main_branch": executor_config.get("main_branch"),
             "run_tests_on_done": post_done.get("run_tests"),
             "run_lint_on_done": post_done.get("run_lint"),