infra/scripts/workstation/claude-hooks/homelab-memory-recall.py
Viktor Barzin 0fa5852ec6
Some checks are pending
Build infra CLI / build (push) Waiting to run
ci/woodpecker/push/default Pipeline was successful
homelab v0.8.2: fix memory recall truncating multibyte UTF-8 mid-character
emo's Claude Code sessions hit "UserPromptSubmit hook error" on almost every
prompt. Root cause: the homelab-memory-recall.py UserPromptSubmit hook runs
`homelab memory recall <prompt>` and strict-decodes its stdout. printMemories
truncated each memory's preview with a BYTE slice (c[:240]), which cuts through
the middle of a 2-byte Cyrillic character and emits invalid UTF-8 (a dangling
0xd0 lead byte). The hook's subprocess.run(text=True) then raised
UnicodeDecodeError — not caught by its `except (TimeoutExpired, OSError)` — so
the hook exited non-zero and Claude surfaced the error. It is Cyrillic-specific
(ASCII has no multibyte chars to split), so it bit emo (Bulgarian prompts) every
turn while English users almost never saw it.

Two-layer fix:
- cli: truncatePreview() now counts RUNES, not bytes, so the preview never
  splits a character. Regression test asserts valid UTF-8 on a long Cyrillic
  string. Fixes the root for every consumer of `memory recall` / `memory list`.
- hook: subprocess.run gains errors="replace" and the except is broadened to
  honor the script's own "best-effort, exit 0" contract — so a truncated or
  otherwise odd payload can never again surface as a hook error.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-28 09:40:51 +00:00

76 lines
2.7 KiB
Python
Executable file

#!/usr/bin/env python3
"""UserPromptSubmit hook: inject relevant memories via `homelab memory recall`.
Replaces the claude-memory MCP recall path. Instead of instructing the model to
call the memory_recall MCP tool, this hook runs the homelab CLI (a direct client
to the same claude-memory HTTP API) and injects the ACTUAL results as context —
so recall is automatic, needs no model tool-call, and works with the MCP
uninstalled. Best-effort: any failure exits 0 silently (recall just doesn't
happen that turn, exactly like the MCP being unavailable).
Wizard-only trial of the MCP deprecation (2026-06-20). Reversible: restore the
plugin command in ~/.claude/settings.json (backup: settings.json.bak-pre-homelab-memory).
"""
import json
import os
import shutil
import subprocess
import sys
def main() -> None:
try:
hook_input = json.load(sys.stdin)
except (json.JSONDecodeError, EOFError):
return
prompt = ""
if isinstance(hook_input, dict):
prompt = hook_input.get("prompt") or hook_input.get("user_prompt") or ""
if not prompt and isinstance(hook_input.get("content"), str):
prompt = hook_input["content"]
prompt = (prompt or "").strip()
# Same gates as the original recall hook: skip short prompts, code/JSON/XML blobs.
if len(prompt) < 10 or prompt[0] in "`{<":
return
homelab = shutil.which("homelab") or "/usr/local/bin/homelab"
if not os.path.exists(homelab):
return
if not (os.environ.get("CLAUDE_MEMORY_API_KEY") or os.environ.get("MEMORY_API_KEY")):
return
try:
res = subprocess.run(
[homelab, "memory", "recall", prompt, "--limit", "5"],
capture_output=True, text=True, errors="replace", timeout=4,
env=os.environ,
)
except Exception:
# Best-effort: ANY failure — timeout, OSError, or a UnicodeDecodeError on
# truncated multibyte (Cyrillic) output — must silently skip recall this
# turn, exactly like the MCP being unavailable. errors="replace" above
# also keeps a mid-rune-truncated payload from raising here at all. Never
# let this hook surface a "UserPromptSubmit hook error".
return
out = (res.stdout or "").strip()
if res.returncode != 0 or not out:
return
context = (
"Relevant stored memories (via `homelab memory recall`) — incorporate "
"naturally if useful; do NOT mention this lookup to the user:\n\n" + out
)
print(json.dumps({
"hookSpecificOutput": {
"hookEventName": "UserPromptSubmit",
"additionalContext": context,
}
}))
if __name__ == "__main__":
main()