hooks/claude-hook.sh answered every event by running resolve-plan-dir.sh and inject-plan.sh, about 130 forks per UserPromptSubmit. Under Git Bash on Windows a fork costs about 90 ms, so a prompt took 7 to 12 s against the 10 s hook timeout and Claude Code discarded the plan context, and PreToolUse added 5 s before every tool call. Measured on the reporting machine: 8.1, 10.2 and 7.1 s per prompt before, 0.33 s after. scripts/inject-plan.py is a byte-identical twin of inject-plan.sh plus the dispatcher logic of claude-hook.sh. The launcher finds a CPython 3 with a fork-free PATH walk, runs the twin with -I -B and falls back to the shell chain on any non-zero exit; PWF_FAST_PATH=0 forces the shell chain; a directory without planning state exits before any fork. The launcher hands the twin its own $PWD spelling so both routes share one turn-marker and one progress-guard slot per plan. skill-hook.sh uses the twin after eligibility and isolates its own two interpreter calls. Both dispatchers run the json_string awk under LC_ALL=C so a character outside the BMP no longer leaves as a lone surrogate on Windows. The twin ships in every scripts directory that ships inject-plan.sh. tests/test_inject_plan_python_parity.py runs both implementations over the same fixtures and asserts identical stdout, including what the second-model review found and this commit fixes: NUL bytes in .active_plan, the awk "phases: /N" quirk for a plan with no completed phase, the CRLF progress tail under Git for Windows sed, cache slots shared across routes, and non-BMP characters through the JSON encoder.
88 lines
3.4 KiB
Python
88 lines
3.4 KiB
Python
"""Location parity for dual-shipped scripts (root scripts/ vs skill scripts/).
|
|
|
|
The hook dispatchers resolve scripts from three locations: the skill install dir
|
|
(``${CLAUDE_SKILL_DIR}/scripts``), the bare-skill known path
|
|
(``~/.claude/skills/planning-with-files/scripts``), and the plugin-marketplace
|
|
fallback (``~/.claude/plugins/marketplaces/planning-with-files/scripts``). The
|
|
marketplace path is the REPO ROOT ``scripts/`` directory, so every script the
|
|
dispatchers call — directly or as a sibling (inject-plan.sh shells
|
|
``${SCRIPT_DIR}/ledger-summary.sh``) — must exist in BOTH locations with
|
|
identical content.
|
|
|
|
This was a live bug before this test existed: the v3 review wave restored
|
|
inject-plan.sh and gate-stop.sh to root scripts/ but not the ledger trio, so the
|
|
marketplace route in autonomous/gated mode silently fell back to injecting the
|
|
raw progress.md tail — the exact surface the structured ledger summary is
|
|
documented to close (SKILL.md "Structured ledger injection").
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import unittest
|
|
from pathlib import Path
|
|
|
|
|
|
REPO_ROOT = Path(__file__).resolve().parents[1]
|
|
ROOT_SCRIPTS = REPO_ROOT / "scripts"
|
|
SKILL_SCRIPTS = REPO_ROOT / "skills" / "planning-with-files" / "scripts"
|
|
|
|
# Every script the hook dispatchers reach from a scripts/ directory, in both
|
|
# shell and PowerShell forms. Sibling dependencies count: inject-plan.sh calls
|
|
# ledger-summary.sh, gate-stop.sh calls check-complete.sh, init-session calls
|
|
# attest-plan and resolve-plan-dir, ledger-summary/phase-status back the
|
|
# autonomous-mode injection and the gate's phase reporting.
|
|
DUAL_SHIPPED = [
|
|
"attest-plan.ps1",
|
|
"attest-plan.sh",
|
|
"check-complete.ps1",
|
|
"check-complete.sh",
|
|
"gate-stop.sh",
|
|
"init-session.ps1",
|
|
"init-session.sh",
|
|
"inject-plan.sh",
|
|
"inject-plan.py",
|
|
"skill-hook.sh",
|
|
"ledger-append.ps1",
|
|
"ledger-append.sh",
|
|
"ledger-summary.ps1",
|
|
"ledger-summary.sh",
|
|
"phase-status.ps1",
|
|
"phase-status.sh",
|
|
"plan-doctor.sh",
|
|
"resolve-plan-dir.ps1",
|
|
"resolve-plan-dir.sh",
|
|
"set-active-plan.ps1",
|
|
"set-active-plan.sh",
|
|
]
|
|
|
|
|
|
class ScriptLocationParityTests(unittest.TestCase):
|
|
def test_dual_shipped_scripts_exist_in_both_locations(self) -> None:
|
|
for name in DUAL_SHIPPED:
|
|
with self.subTest(script=name):
|
|
self.assertTrue(
|
|
(ROOT_SCRIPTS / name).is_file(),
|
|
f"scripts/{name} missing — plugin-marketplace fallback route breaks",
|
|
)
|
|
self.assertTrue(
|
|
(SKILL_SCRIPTS / name).is_file(),
|
|
f"skills/planning-with-files/scripts/{name} missing — skill install route breaks",
|
|
)
|
|
|
|
def test_dual_shipped_scripts_are_byte_identical(self) -> None:
|
|
for name in DUAL_SHIPPED:
|
|
root_file = ROOT_SCRIPTS / name
|
|
skill_file = SKILL_SCRIPTS / name
|
|
if not (root_file.is_file() and skill_file.is_file()):
|
|
continue # existence failures are reported by the test above
|
|
with self.subTest(script=name):
|
|
self.assertEqual(
|
|
root_file.read_bytes(),
|
|
skill_file.read_bytes(),
|
|
f"{name} drifted between scripts/ and skill scripts/ — "
|
|
"the two install routes would behave differently",
|
|
)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
unittest.main()
|