feat(auto_heal): multi-profile key switching + fix two error classes

Error investigation (from restored monitoring) found 2 root causes:

1. 'Blocked: script path resolves outside scripts dir' (5+ cron jobs):
   Jul 17 symlink refactor replaced real scripts with symlinks; the hermes
   cron scheduler's security check (Path.resolve + relative_to) rejects
   symlink escape. ALL no_agent script jobs blocked since Jul 17 23:12.
   FIX: converted 102 symlinks to hardlinks (same inode, resolve() stays
   inside scripts_dir, single-source still works). Permanent structural fix.

2. HTTP 429 on default profile (知识研究/梦境循环/wiki-self-growth/
   evolution-pulse/大脑任务执行): default gateway used ocg-key1 (weekly
   100%). FIX: switched default profile to ocg-key6 + added missing
   provider block. LLM verified working (3.1s).

auto_heal extended to actually cover these automatically next time:
- PROFILES registry: zhiwei (8643, system svc) + default (8642, user svc)
- current_provider/switch_key/_scan_agent_log parameterized by profile
- health() now reports llm_provider_default (agent.log scan)
- auto_heal: per-profile 429 detection -> best_key -> switch_key(profile)
- _ensure_provider_block: injects missing provider credentials from
  zhiwei config (single source of truth) into target config
This commit is contained in:
hmo
2026-07-20 01:12:38 +08:00
parent a05c118cbc
commit 57377e9dd3
8 changed files with 338 additions and 73 deletions
+33
View File
@@ -0,0 +1,33 @@
import os, sys
scripts_dir = '/home/hmo/.hermes/profiles/position-analyst/scripts'
fixed, skipped, errors = 0, 0, []
for name in os.listdir(scripts_dir):
p = os.path.join(scripts_dir, name)
if not os.path.islink(p):
skipped += 1
continue
target = os.path.realpath(p)
if not os.path.exists(target):
errors.append(f"{name}: broken symlink -> {target}")
continue
try:
os.unlink(p)
os.link(target, p) # hardlink: same inode, .resolve() stays in scripts_dir
fixed += 1
except Exception as e:
errors.append(f"{name}: {e}")
print(f"fixed: {fixed}, skipped (non-symlink): {skipped}, errors: {len(errors)}")
for e in errors:
print(' ', e)
# verify with the scheduler's own check
from pathlib import Path
scripts_resolved = Path(scripts_dir).resolve()
test = (Path(scripts_dir) / 'meta_growth.py').resolve()
try:
test.relative_to(scripts_resolved)
print('VERIFY OK: meta_growth.py now passes relative_to check')
except ValueError:
print('VERIFY FAIL: still resolves outside')