agentic-devtools 0.2.131__tar.gz → 0.2.133__tar.gz

agentic_devtools-0.2.133/.github/agents/copilot-instructions.md

@@ -0,0 +1,74 @@
+# agentic-devtools Development Guidelines
+
+Auto-generated from all feature plans. Last updated: 2026-02-03
+
+## Active Technologies
+
+- Markdown (repository documentation) + None (001-separate-docs)
+- Markdown (YAML frontmatter) for agents/prompts; Python 3.11 for optional coverage script + VS Code Copilot Chat (agent `.agent.md` and prompt `.prompt.md` conventions) (001-add-workflow-step-agents)
+- N/A — all artifacts are static Markdown files committed to the repository (001-add-workflow-step-agents)
+
+- Python >= 3.8 + requests, Jinja2 (bestehend), neu: build, twine (für Release-Flows) (001-pypi-wheel-release)
+
+## Project Structure
+
+```text
+src/
+tests/
+```
+
+## Commands
+
+```bash
+cd src
+pytest
+ruff check .
+```
+
+## Code Style
+
+Python >= 3.8: Follow standard conventions
+
+## Recent Changes
+
+- 001-add-workflow-step-agents: Added Markdown (YAML frontmatter) for agents/prompts; Python 3.11 for optional coverage script + VS Code Copilot Chat (agent `.agent.md` and prompt `.prompt.md` conventions)
+- 001-separate-docs: Added Markdown (repository documentation) + None
+
+- 001-pypi-wheel-release: Added Python >= 3.8 + requests, Jinja2 (bestehend), neu: build, twine (für Release-Flows)
+
+<!-- MANUAL ADDITIONS START -->
+
+## Pre-Push Hooks
+
+When git hooks are enabled (`core.hooksPath=.githooks`), pre-push hooks enforce
+the following checks before each push:
+
+- `ruff format` — code formatting (not black)
+- `ruff check` — linting and import sorting
+- Per-file 100% test coverage
+- `mypy` — static type checking
+- Test structure validation
+
+Markdown style is validated in CI via markdownlint on changed `.md` files.
+
+Agents working on this repo do **not** need to manually run `ruff` or `mypy` before
+pushing when hooks are enabled — the pre-push hook handles it automatically.
+If push is rejected, fix the reported issues, run `agdt-git-save-work`, and retry.
+
+## CI Pipeline (4-Tier Gates)
+
+The CI pipeline uses the following gate names:
+
+1. `Targeted Checks ✅` — ruff/mypy/test-structure + per-file coverage on changed files (plus markdownlint for changed `.md`)
+2. `Smart Module Tests ✅` — targeted test execution for affected modules
+3. `Workflow Tests ✅` — workflow integration tests for workflow-related changes
+4. `Copilot Review ✅` — automated AI code review
+
+All four gates must pass for a PR to be mergeable.
+
+## Formatting
+
+- Use `ruff format .` as the formatter (not black)
+- Use `ruff check .` for linting (not flake8 or isort separately)
+
+<!-- MANUAL ADDITIONS END -->

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/.github/agents/senior-python-developer.md

@@ -542,7 +542,7 @@ When reviewing or writing code, prioritize:
 - **ORM**: SQLAlchemy 2.0+, Django ORM, or Tortoise ORM
 - **Testing**: pytest, pytest-asyncio, pytest-cov
 - **Type Checking**: mypy, pyright
-- **Linting**: ruff (or black + flake8 + isort)
+- **Linting & Formatting**: ruff (handles linting, formatting, and import sorting)
 - **Dependency Management**: poetry or pip-tools
 - **API Framework**: FastAPI, Django REST Framework
 - **Async**: asyncio, aiohttp for async operations
@@ -550,23 +550,36 @@ When reviewing or writing code, prioritize:
 ### Code Quality Tools
 
 ```bash
-# Format code
-black .
-isort .
+# Format code (replaces black + isort)
+ruff format .
 
-# Type check
-mypy src/
-
-# Lint
+# Lint (replaces flake8 + isort)
 ruff check .
 
-# Test with coverage
-pytest --cov=src --cov-report=html
+# Type check
+mypy .
+
+# Run tests
+agdt-test
 
 # Security check
-bandit -r src/
+bandit -r agentic_devtools/
 ```
 
+### Pre-Push Hook Automation
+
+When git hooks are enabled (`core.hooksPath=.githooks`), pre-push hooks enforce
+the following before each push:
+
+- `ruff format` — code formatting
+- `ruff check` — linting
+- Per-file 100% test coverage
+- `mypy` — type checking
+- Test structure validation
+
+If push is rejected by the pre-push hook, fix the reported issues, run `agdt-git-save-work`,
+and retry the push.
+
 ## Communication Style
 
 - **Professional**: Clear, precise technical communication

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/.github/prompts/agdt.address-copilot-review.ci-repair.prompt.md

@@ -135,8 +135,20 @@ Document a triage table:
 If CI failure context is provided (see the `## CI Failure Context` section at the
 end of this prompt, if present), address those failures:
 
+The CI pipeline gates are:
+
+- `Targeted Checks ✅` — ruff/mypy/test-structure + per-file coverage on changed files (plus markdownlint for changed `.md`)
+- `Smart Module Tests ✅` — targeted test execution for affected modules
+- `Workflow Tests ✅` — workflow integration tests for workflow-related changes
+- `Copilot Review ✅` — automated AI code review
+
 ### For Lint Failures (ruff)
 
+> **Note:** Lint/format failures reaching CI are uncommon — the pre-push hook
+> enforces `ruff format` and `ruff check` locally. If these failures appear in
+> CI, the hook was likely bypassed with `--no-verify` or not enabled
+> (`core.hooksPath` not set to `.githooks`).
+
 1. Read the failure messages to identify specific files and violations
 2. Apply automated fixes using **only pinned trusted tooling**:
 

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/.github/prompts/agdt.address-copilot-review.evaluate-and-respond.prompt.md

@@ -188,8 +188,20 @@ Assign a confidence level for each suggestion:
 If the `## CI Failures` section is present in the trigger comment, address those failures
 before making code changes for review comments.
 
+The CI pipeline gates are:
+
+- `Targeted Checks ✅` — ruff/mypy/test-structure + per-file coverage on changed files (plus markdownlint for changed `.md`)
+- `Smart Module Tests ✅` — targeted test execution for affected modules
+- `Workflow Tests ✅` — workflow integration tests for workflow-related changes
+- `Copilot Review ✅` — automated AI code review
+
 ### For Lint Failures (ruff)
 
+> **Note:** Lint/format failures reaching CI are uncommon — the pre-push hook
+> enforces `ruff format` and `ruff check` locally. If these failures appear in
+> CI, the hook was likely bypassed with `--no-verify` or not enabled
+> (`core.hooksPath` not set to `.githooks`).
+
 Run Ruff only on Python files already in the PR diff:
 
 ```bash

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentic-devtools
-Version: 0.2.131
+Version: 0.2.133
 Summary: Agentic devtools integrate Jira, DevOps & more
 Author: ayaiayorg
 License-Expression: MIT

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/_version.py

@@ -18,7 +18,7 @@ version_tuple: tuple[int | str, ...]
 commit_id: str | None
 __commit_id__: str | None
 
-__version__ = version = '0.2.131'
-__version_tuple__ = version_tuple = (0, 2, 131)
+__version__ = version = '0.2.133'
+__version_tuple__ = version_tuple = (0, 2, 133)
 
 __commit_id__ = commit_id = None

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/actionable_checks.py

@@ -4,9 +4,9 @@ from __future__ import annotations
 
 DEFAULT_ACTIONABLE_CHECK_NAMES = frozenset(
     {
-        "Targeted Checks ✅",
-        "Smart Module Tests ✅",
-        "Workflow Tests ✅",
+        "Run Targeted Checks",
+        "Run Smart Module Tests",
+        "Workflow Tests",
         "Code scanning results / CodeQL",
         "CodeQL / Analyze (actions) (dynamic)",
         "CodeQL / Analyze (python) (dynamic)",

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/pipeline/actions/dispatch_repair.py

@@ -33,12 +33,11 @@ class DispatchRepairAction:
     """Dispatch a repair when CI fails or actionable Copilot review exists.
 
     Preconditions:
-    - No active Copilot session
     - CI failed OR actionable Copilot review on HEAD
     - Deduplication limit not exceeded
     - Cycle limit not exceeded
 
-    Idempotency: Active session or recent dispatch → skip.
+    Idempotency: Recent dispatch → skip.
     """
 
     @property
@@ -49,16 +48,6 @@ class DispatchRepairAction:
         """Evaluate whether repair dispatch is needed."""
         preconditions: dict[str, bool] = {}
 
-        # No active session
-        preconditions["no_active_session"] = not snapshot.active_session
-        if snapshot.active_session:
-            return ActionResult(
-                name=self.name,
-                decision=ActionDecision.SKIP,
-                preconditions=preconditions,
-                details="Copilot coding session is active",
-            )
-
         # CI failed OR actionable review
         ci_failing = snapshot.ci_status == "failing"
         review_actionable = _is_copilot_review_actionable(snapshot)

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/pipeline/actions/request_review.py

@@ -6,6 +6,7 @@ import logging
 
 from agentic_devtools.cli.ci.models import COPILOT_REVIEWER_LOGIN
 from agentic_devtools.cli.ci.pipeline.models import ActionDecision, ActionResult
+from agentic_devtools.cli.ci.pipeline.session_detector import is_copilot_session_active_via_agent_task
 from agentic_devtools.cli.ci.pipeline.snapshot import DerivedState, PRStateSnapshot
 from agentic_devtools.cli.ci.provider import CIPlatformProvider
 
@@ -20,11 +21,11 @@ class RequestReviewAction:
     Preconditions:
     - PR is not draft (uses DerivedState)
     - Repair was not dispatched in this pipeline run
-    - No active Copilot coding session
     - No unresolved review threads
     - CI is passing
     - No effective Copilot review on HEAD
     - Copilot not already requested as reviewer
+    - No active Copilot coding session
 
     Idempotency: Review exists or pending → skip.
     """
@@ -59,16 +60,6 @@ class RequestReviewAction:
                 details="Repair dispatched — deferring review request",
             )
 
-        # Guard: never request review when active coding session is in progress
-        preconditions["no_active_session"] = not snapshot.active_session
-        if snapshot.active_session:
-            return ActionResult(
-                name=self.name,
-                decision=ActionDecision.SKIP,
-                preconditions=preconditions,
-                details="Copilot session active — deferring review request",
-            )
-
         # Guard: block review request when unresolved review threads exist
         unresolved = derived.unresolved_threads
         preconditions["no_unresolved_threads"] = unresolved == 0
@@ -113,6 +104,17 @@ class RequestReviewAction:
                 details="Copilot review already requested",
             )
 
+        # Guard: never request review when active coding session is in progress
+        active_session = is_copilot_session_active_via_agent_task(snapshot.base_repo_full_name, snapshot.pr_number)
+        preconditions["no_active_session"] = not active_session
+        if active_session:
+            return ActionResult(
+                name=self.name,
+                decision=ActionDecision.SKIP,
+                preconditions=preconditions,
+                details="Copilot session active — deferring review request",
+            )
+
         return ActionResult(
             name=self.name,
             decision=ActionDecision.EXECUTE,

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/pipeline/actions/squash.py

@@ -5,6 +5,7 @@ from __future__ import annotations
 import logging
 
 from agentic_devtools.cli.ci.pipeline.models import ActionDecision, ActionResult
+from agentic_devtools.cli.ci.pipeline.session_detector import is_copilot_session_active_via_agent_task
 from agentic_devtools.cli.ci.pipeline.snapshot import DerivedState, PRStateSnapshot
 from agentic_devtools.cli.ci.provider import CIPlatformProvider
 
@@ -17,8 +18,8 @@ class SquashAction:
     Preconditions:
     - Commits above merge-base > 1
     - No repair dispatched in this run (keeps HEAD stable for the repair cycle)
-    - No active Copilot coding session (pending review does NOT block squash)
     - CI passing
+    - No active Copilot coding session (pending review does NOT block squash)
 
     Idempotency: Already 1 commit → skip.
     """
@@ -52,24 +53,25 @@ class SquashAction:
                 details="Repair dispatched — deferring squash",
             )
 
-        # No active Copilot session (coding/repair only — pending review does NOT block squash)
-        preconditions["no_active_session"] = not snapshot.active_session
-        if snapshot.active_session:
+        # CI must be passing
+        preconditions["ci_passing"] = snapshot.ci_status == "passing"
+        if snapshot.ci_status != "passing":
             return ActionResult(
                 name=self.name,
                 decision=ActionDecision.SKIP,
                 preconditions=preconditions,
-                details="Copilot session active — deferring squash",
+                details=f"CI is {snapshot.ci_status} — deferring squash",
             )
 
-        # CI must be passing
-        preconditions["ci_passing"] = snapshot.ci_status == "passing"
-        if snapshot.ci_status != "passing":
+        # No active Copilot session (coding/repair only — pending review does NOT block squash)
+        active_session = is_copilot_session_active_via_agent_task(snapshot.base_repo_full_name, snapshot.pr_number)
+        preconditions["no_active_session"] = not active_session
+        if active_session:
             return ActionResult(
                 name=self.name,
                 decision=ActionDecision.SKIP,
                 preconditions=preconditions,
-                details=f"CI is {snapshot.ci_status} — deferring squash",
+                details="Copilot session active — deferring squash",
             )
 
         return ActionResult(

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/pipeline/session_detector.py

@@ -1,13 +1,19 @@
 """Copilot session activity detector.
 
-Replaces the squash-wait state machine with a simple check of the
-Issues Events API.
+Provides two detection strategies:
+- ``is_copilot_session_active_via_agent_task``: Uses ``gh agent-task list`` for
+  authoritative session state (preferred, fail-open).
+- ``is_copilot_session_active``: Legacy events-based heuristic (deprecated,
+  fail-closed).
 """
 
 from __future__ import annotations
 
+import json
 import logging
 import os
+import subprocess
+import warnings
 from datetime import datetime, timezone
 
 from agentic_devtools.cli.ci.models import (
@@ -16,9 +22,13 @@ from agentic_devtools.cli.ci.models import (
     COPILOT_SESSION_EVENT_STARTED,
 )
 from agentic_devtools.cli.ci.provider import CIPlatformProvider
+from agentic_devtools.cli.subprocess_utils import run_safe
 
 logger = logging.getLogger(__name__)
 
+_ACTIVE_TASK_STATUSES = frozenset({"queued", "requested", "waiting", "in_progress", "running"})
+_DEFAULT_AGENT_TASK_TIMEOUT_SECONDS = 10
+
 _DEFAULT_MAX_SESSION_AGE_SECONDS = 3600  # 1 hour
 
 
@@ -61,18 +71,116 @@ def _is_session_stale(created_at: str, max_age_seconds: int) -> bool:
     return age > max_age_seconds
 
 
+def is_copilot_session_active_via_agent_task(
+    repo: str,
+    pr_number: int,
+    *,
+    timeout_seconds: int = _DEFAULT_AGENT_TASK_TIMEOUT_SECONDS,
+) -> bool:
+    """Check if a Copilot coding session is active using ``gh agent-task list``.
+
+    This is the preferred session detector. It queries the GitHub CLI for
+    authoritative agent task state and uses **fail-open** semantics: if the
+    command fails for any reason the function returns ``False`` (no active
+    session) rather than blocking automation.
+
+    Args:
+        repo: Full repository name (e.g. ``owner/repo``).
+        pr_number: Pull request number to check.
+        timeout_seconds: Maximum time to wait for the ``gh`` subprocess.
+
+    Returns:
+        True if at least one agent task for the given PR is in an active status
+        (queued, requested, waiting, in_progress, running).
+        False otherwise or on any error (fail-open).
+    """
+    cmd = [
+        "gh",
+        "agent-task",
+        "list",
+        "--repo",
+        repo,
+        "--json",
+        "id,status,pullRequestNumber,createdAt",
+    ]
+    try:
+        result = run_safe(
+            cmd,
+            capture_output=True,
+            text=True,
+            shell=False,
+            timeout=timeout_seconds,
+        )
+    except subprocess.TimeoutExpired:
+        logger.warning(
+            "PR #%d: gh agent-task list timed out after %ds — assuming no active session (fail-open)",
+            pr_number,
+            timeout_seconds,
+        )
+        return False
+    except (OSError, FileNotFoundError, PermissionError) as exc:
+        logger.warning(
+            "PR #%d: gh agent-task list failed — assuming no active session (fail-open): %s",
+            pr_number,
+            exc,
+        )
+        return False
+
+    if result.returncode != 0:
+        logger.warning(
+            "PR #%d: gh agent-task list exited with code %d — assuming no active session (fail-open)",
+            pr_number,
+            result.returncode,
+        )
+        return False
+
+    try:
+        tasks = json.loads(result.stdout)
+    except (json.JSONDecodeError, ValueError) as exc:
+        logger.warning(
+            "PR #%d: gh agent-task list returned malformed JSON — assuming no active session (fail-open): %s",
+            pr_number,
+            exc,
+        )
+        return False
+
+    if not isinstance(tasks, list):
+        logger.warning(
+            "PR #%d: gh agent-task list returned non-list JSON — assuming no active session (fail-open)",
+            pr_number,
+        )
+        return False
+
+    for task in tasks:
+        if not isinstance(task, dict):
+            continue
+        task_pr = task.get("pullRequestNumber")
+        if task_pr == pr_number and task.get("status") in _ACTIVE_TASK_STATUSES:
+            logger.info(
+                "PR #%d: Active agent task detected (id=%s, status=%s)",
+                pr_number,
+                task.get("id"),
+                task.get("status"),
+            )
+            return True
+
+    return False
+
+
 def is_copilot_session_active(provider: CIPlatformProvider, pr_number: int) -> bool:
     """Check if a Copilot coding session is currently active.
 
+    .. deprecated::
+        Use :func:`is_copilot_session_active_via_agent_task` instead.
+        This function uses an unreliable events-based heuristic and will be
+        removed in a future release.
+
     Looks for the latest copilot_work_started event and checks whether
     a terminal event (finished or failure) exists with a higher ID.
     Additionally applies a staleness timeout: if the latest start event
     is older than ``AGDT_MAX_SESSION_AGE_SECONDS`` (default 3600s) with
     no terminal event, the session is considered stale/inactive.
 
-    This single check replaces the entire squash-wait state machine for
-    actions 4 (resolve threads), 5 (dispatch repair), and 6 (squash).
-
     Args:
         provider: CI platform provider.
         pr_number: Pull request number.
@@ -83,6 +191,11 @@ def is_copilot_session_active(provider: CIPlatformProvider, pr_number: int) -> b
         False when a terminal event is confirmed after the latest start,
         or when the latest start event exceeds the staleness threshold.
     """
+    warnings.warn(
+        "is_copilot_session_active() is deprecated. Use is_copilot_session_active_via_agent_task() instead.",
+        DeprecationWarning,
+        stacklevel=2,
+    )
     try:
         events = provider.list_pr_issue_events(pr_number)
     except Exception as exc:

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/pipeline/snapshot.py

@@ -11,7 +11,6 @@ from agentic_devtools.cli.ci.models import (
     CheckRunStatus,
     ReviewInfo,
 )
-from agentic_devtools.cli.ci.pipeline.session_detector import is_copilot_session_active
 from agentic_devtools.cli.ci.provider import CIPlatformProvider
 
 
@@ -170,9 +169,6 @@ def build_pr_state_snapshot(
     # Check Copilot review pending
     copilot_review_pending = _is_copilot_review_pending(pr_meta.requested_reviewers)
 
-    # Check active Copilot session
-    active_session = is_copilot_session_active(provider, pr_number)
-
     # Count commits above merge-base
     commit_count = _count_commits(provider, base_branch=pr_meta.base_branch, head_sha=pr_meta.head_sha)
 
@@ -187,7 +183,6 @@ def build_pr_state_snapshot(
         review_state=copilot_review_state,
         copilot_review_id=copilot_review_id,
         copilot_review_inline_count=copilot_review_inline_count,
-        active_session=active_session,
         copilot_review_pending=copilot_review_pending,
         unresolved_threads=unresolved_threads,
         labels=list(pr_meta.labels),

{agentic_devtools-0.2.131 → agentic_devtools-0.2.133}/agentic_devtools/cli/ci/pipeline/summary.py

@@ -164,7 +164,7 @@ def _render_state_snapshot(snapshot: PRStateSnapshot) -> str:
     lines = [
         f"- HEAD: `{head_sha_short}`",
         f"- Commits above merge-base: {_sanitize_cell(str(snapshot.commit_count))}",
-        f"- Copilot session active: {_sanitize_cell(str(snapshot.active_session))}",
+        "- Copilot session active: N/A",
         f"- Copilot review on HEAD: {review_state}{inline_suffix}",
         f"- CI: {_sanitize_cell(snapshot.ci_status)}",
         f"- Unresolved threads: {_sanitize_cell(str(snapshot.unresolved_threads))}",

agentic_devtools-0.2.133/specs/1643-remove-active-session-gate/analysis-report.md

@@ -0,0 +1,59 @@
+# Cross-Artifact Consistency & Quality Analysis Report
+
+## Findings Table
+
+| ID | Category | Severity | Location(s) | Summary | Recommendation |
+|----|----------|----------|-------------|---------|----------------|
+| F-01 | B | MEDIUM | Spec FR-003 | "GitHub CLI ≥ version TBD" in plan Dependencies section — unresolved placeholder | Specify minimum `gh` CLI version that supports `agent-task` subcommand or document that fail-open handles older versions gracefully |
+| F-02 | B | LOW | Spec FR-005 | Active status set `{"queued","requested","waiting","in_progress","running"}` stated without citing authoritative source | Add a reference to GitHub documentation or note this is empirically determined; spec is otherwise clear |
+| F-03 | C | MEDIUM | Plan Phase 3, Task T016 | "Remove lines 52–60" is brittle line-number reference; code may shift before implementation | Reference by code pattern (`no_active_session` precondition block) rather than line numbers |
+| F-04 | C | LOW | Spec FR-003 | Optional `gh agent-task view <id>` compatibility fallback mentioned but no test coverage task exists | Add test task for fallback path or explicitly state it is not implemented in this iteration |
+| F-05 | E | LOW | NFR-001 | NFR-001 (10s timeout) has no dedicated task — coverage is implicit via T005/T007 | Acceptable since T005 explicitly tests timeout; note for traceability |
+| F-06 | E | LOW | NFR-002 | NFR-002 (no new dependencies) has no explicit verification task | Add a grep-based verification step to T039 or Phase 8 |
+| F-07 | E | LOW | NFR-004 | NFR-004 (logging patterns) has no dedicated verification task | Add assertion in T007/T008 tests that WARNING logs use module-level logger |
+| F-08 | F | LOW | Spec FR-003 vs Plan Phase 1 | Spec says `--json id,status,pullRequestNumber,createdAt`; plan repeats verbatim — consistent but worth noting any future schema drift risk | No action needed; consistency confirmed |
+| F-09 | F | MEDIUM | Tasks T024/T025 | Tasks marked `[P]` (parallel?) but have dependency on T009 AND T018 — cannot start until Phase 4 completes | Clarify `[P]` annotation meaning or remove if not parallel-safe |
+| F-10 | D | LOW | Spec | No explicit "Out of Scope" section listing what is NOT being changed | Add brief out-of-scope section (e.g., no changes to `ResolveThreadsAction`, no new pipeline commands) |
+
+### Category G Structured Findings
+
+[]
+
+## Coverage Summary Table
+
+| Requirement Key | Has Task? | Task IDs | Notes |
+|----------------|-----------|----------|-------|
+| FR-001 | ✅ | T013, T014, T016, T017 | Full coverage |
+| FR-002 | ✅ | T013, T015, T016 | Full coverage |
+| FR-003 | ✅ | T002, T003, T007 | Full coverage |
+| FR-004 | ✅ | T005, T006, T008 | Full coverage |
+| FR-005 | ✅ | T002, T003, T004, T007 | Full coverage |
+| FR-006 | ✅ | T010, T011, T012 | Full coverage |
+| FR-007 | ✅ | T019, T020, T021, T022, T023, T040 | Full coverage |
+| FR-008 | ✅ | T024–T035, T039, T040 | Full coverage |
+| NFR-001 | ✅ (implicit) | T005, T007 | Timeout tested in T005; implemented in T007 |
+| NFR-002 | ⚠️ | T039 (partial) | No explicit dependency-audit task |
+| NFR-003 | ✅ | T036, T037 | Coverage enforced by targeted-checks |
+| NFR-004 | ✅ (implicit) | T007, T008 | Logging implemented alongside error handling |
+
+## Metrics
+
+| Metric | Value |
+|--------|-------|
+| Total Requirements | 12 (8 FR + 4 NFR) |
+| Total Tasks | 40 |
+| Coverage % | 100% (FR), 92% (NFR — NFR-002 implicit only) |
+| Ambiguity Count | 2 (F-01, F-02) |
+| Requirement Duplication Count (Category A) | 0 |
+| Critical Issues Count | 0 |
+| Task Deduplication Finding Count | 0 |
+| Task Deduplication by Type | duplicate: 0 / overlapping: 0 / conflicting: 0 |
+| Multi-Task Group Count | 0 |
+
+## Next Actions
+
+- Resolve or explicitly accept findings F-01 through F-07 (especially the `gh` minimum version decision and the NFR-002 verification step).
+- Would you like suggested concrete remediation edits for the top findings above?
+
+---
+*Generated by Copilot SDK (claude-opus-4.6)*

agentic_devtools-0.2.133/specs/1643-remove-active-session-gate/fr-coverage.json

@@ -0,0 +1 @@
+{"covered": ["FR-001", "FR-002", "FR-003", "FR-004", "FR-005", "FR-006", "FR-007", "FR-008"], "uncovered": [], "total": 8, "max_retries": 2}