claude-dev-env 8.96.5 → 8.96.7

package/.agents/skills/pr-lifecycle/SKILL.md

@@ -123,7 +123,7 @@ An image a PR needs as visual evidence is not an exception to that row. Upload i
 3. Fix systematically, marking each todo complete.
 4. Reply to each comment inline.
 
-Repair only the reported findings.
+Repair only the reported findings. A finding that comes back on a second head escalates. Fix the source the reviewer reads, or change the gate that files the finding. Post no second reply with the same text, because a repeat finding shows the first reply changed nothing the reviewer sees.
 
 Every `gh` post in this workflow uses `--body-file` per [gh-cli-conventions.md](#gh-cli-conventions) and keeps volatile scratch paths out per [durable-post-artifacts.md](#github-post-input-rules). Stage session edits per [re-stage-before-commit.md](#re-stage-session-edits-before-commit) before each commit.
 
@@ -458,7 +458,7 @@ slurped array and take the last element. For one page, cap the request with a
 
 Stage the files you edited this session right before you commit them. A plain `git commit` records only the staged snapshot; a tracked file this session changed but left unstaged stays behind in the working tree.
 
-No hook denies a commit that would drop tracked session edits. Run `git status` before you commit, then stage what you changed with `git add <paths>` or commit with `git commit -a`. Before you say nothing is left to commit, run `git status` in every worktree the session wrote to, as `git worktree list` lists them.
+No hook denies a commit that would drop tracked session edits. Run `git status` before you commit, then stage what you changed with `git add <paths>` or commit with `git commit -a`. Before you say nothing is left to commit, run `git status` in every worktree the session wrote to, as `git worktree list` lists them. On PowerShell, write the commit message to a scratch file and run `git commit -F <file>`. A here-string piped into `git commit -F -` makes git read the message lines as pathspecs, and the commit fails.
 
 Staging covers tracked files you edited. Do not commit untracked files unless the user explicitly instructs it. An untracked file in the working tree is outside the change until they say otherwise.
 

package/.agents/skills/sync-dev-env/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: sync-dev-env
-description: Install or update claude-dev-env in a running session, link the skills of every enabled plugin so a running session loads them, and schedule a daily re-sync at 6:00 AM local by default. Use when the user asks to sync claude-dev-env into a session, to load the skills of a plugin without a restart, to set up the daily sync, to check which session prompt a session uses, or to turn the claude-dev-env session prompt off for one session.
+description: Install or update claude-dev-env in a running session, link the skills of every enabled plugin so a running session loads them, and schedule a daily re-sync at 6:00 AM local by default. Use when the user asks to sync claude-dev-env into a session, to load the skills of a plugin without a restart, when a /skill fails to resolve, to set up the daily sync, to check which session prompt a session uses, or to turn the claude-dev-env session prompt off for one session.
 argument-hint: "[HH:MM] [IANA time zone]"
 ---
 
@@ -8,6 +8,10 @@ argument-hint: "[HH:MM] [IANA time zone]"
 
 An always-on session starts once, so a SessionStart hook runs once and never again. Claude Code reads plugins only at session start, and a running session has no plugin reload. The skills directory is read live, so one symlink per skill makes each skill of a plugin load at once.
 
+## A /skill fails to resolve
+
+Run `/reload-skills` first. It reloads the skills directory in the running session, so the session keeps its context. When the skill still fails to resolve, run the linker in step 2.
+
 ## 1. Install or update the package
 
 Run `npx --yes claude-dev-env@latest --update --target ~/.claude`.

package/hooks/blocking/test_final_reply_gate.py

@@ -87,6 +87,23 @@ def test_should_allow_a_run_a_tool_result_shows(
     assert run_gate(monkeypatch, capsys, MEASURED_REPLY, transcript_path) == (0, "")
 
 
+def test_should_steer_a_skipped_check_to_an_open_line(
+    monkeypatch: pytest.MonkeyPatch, capsys: pytest.CaptureFixture[str]
+) -> None:
+    skipped_check_reply = "I have not checked the pivot, reach, or screen 26 captures yet."
+    _, stdout_text = run_gate(monkeypatch, capsys, skipped_check_reply)
+    block_reason = json.loads(stdout_text)["reason"]
+    assert "`Open: <check> - <next step>`" in block_reason
+    assert "leave the claim out" not in block_reason
+
+
+def test_should_allow_an_open_line_for_a_skipped_check(
+    monkeypatch: pytest.MonkeyPatch, capsys: pytest.CaptureFixture[str]
+) -> None:
+    open_line = "Open: pivot, reach, and screen 26 checks - the next check-in measures them."
+    assert run_gate(monkeypatch, capsys, open_line) == (0, "")
+
+
 def test_should_allow_a_closing_reply_with_no_message(
     monkeypatch: pytest.MonkeyPatch, capsys: pytest.CaptureFixture[str]
 ) -> None:

package/hooks/blocking/test_reply_length_gate.py

@@ -1000,8 +1000,8 @@ def test_should_collect_each_question_card_prose_field() -> None:
 def test_should_quote_the_hedged_sentence() -> None:
     assert reply_length_gate.unvalidated_claim_violation("Done. It should take a few minutes.") == (
         'No tool validated this: "It should take a few minutes.". Run the check yourself or through'
-        " a subagent, then state what the run, log line, output, or file and line shows, or leave"
-        " the claim out."
+        " a subagent, then state what the run, log line, output, or file and line shows. When you"
+        " cannot run it this turn, report it as `Open: <check> - <next step>`."
     )
 
 

package/hooks/hooks_constants/reply_length_gate_constants.py

@@ -147,7 +147,8 @@ DURATION_CLAIM_PATTERN = re.compile(
 )
 UNVALIDATED_CLAIM_MESSAGE = (
     'No tool validated this: "{claim}". Run the check yourself or through a subagent,'
-    " then state what the run, log line, output, or file and line shows, or leave the claim out."
+    " then state what the run, log line, output, or file and line shows. When you cannot run it"
+    " this turn, report it as `Open: <check> - <next step>`."
 )
 CITED_NUMBER_PATTERN = re.compile(
     r"/actions/runs/(\d+)|/pull/(\d+)|\b(?:run|job|check)\s+#?(\d{5,})\b", re.IGNORECASE

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "claude-dev-env",
-    "version": "8.96.5",
+    "version": "8.96.7",
     "description": "Claude Code development standards — rules, hooks, agents, commands, and skills",
     "type": "module",
     "bin": {