hermes-config-troubleshooting
Diagnose and fix Hermes Agent configuration file corruption, especially auth.json.
- License 许可
- MIT
- 所属仓库
- zapabob/Skills
- 最近更新
- 2026-10-26
- Stars
- ★ 2
- Forks
- 0
01 前言 Front Matter
- name
- hermes-config-troubleshooting
- author
- Hermes Agent
- license
- MIT
- version
- 0.1.0
- platforms
- [ "windows", "macos", "linux" ]
- description
- Diagnose and fix Hermes Agent configuration file corruption, especially auth.json.
- metadata.hermes.tags
- hermes, config, troubleshooting, auth.json
- metadata.hermes.category
- devops
- metadata.hermes.related_skills
- hermes-agent, computer-use
03 SKILL.md 正文
Hermes Config Troubleshooting Skill
When to Use
Use this skill when you suspect Hermes configuration files (e.g., auth.json, config.yaml) are corrupted, causing authentication errors, tool failures, or unexpected behavior. Typical signs include JSON parsing errors, missing credentials, or Hermes failing to start.
Prerequisites
- Hermes Agent installed and accessible via CLI.
- Basic familiarity with JSON and YAML formats.
- Access to the Hermes home directory (
~/.hermesor%USERPROFILE%\.hermes).
How to Run
- Invoke the skill via
hermes skills run hermes-config-troubleshooting(or use the agent directly). - Follow the interactive prompts or execute the steps outlined in the Procedure section.
- Optionally, reference the detailed notes in
references/auth-json-corruption.md.
Quick Reference
- Check JSON validity:
python -m json.tool <file> - View file:
hermes read_file --path <file> - Backup before editing: copy the file to
.bak - Restore from backup: copy
.bakback to original location - Re-authenticate: run
hermes authto refresh tokens
Procedure
-
Locate the suspect file The auth.json lives in your active Hermes home directory. Run
echo $HERMES_HOME(or checkhermes config show paths)- default is
~/.hermes/, profiles use~/.hermes/profiles/<name>/. Typical path:$HERMES_HOME/auth.json.
- default is
-
Validate JSON syntax Run:
python -m json.tool /path/to/auth.jsonIf it returns an error, note the line and column indicated.
-
Inspect the file content Use
hermes read_fileor a text editor to view the problematic area. Common issues:- Truncated strings (missing closing quotes)
- Trailing commas
- Invalid control characters
- Duplicate keys
-
Create a backup Copy the file to
auth.json.bakbefore making changes. -
Fix the JSON
- Remove trailing commas.
- Ensure all strings are properly quoted and escaped.
- Validate that the structure matches the expected schema (top‑level object with
active_provider,credential_pool,providers,updated_at,version). - If unsure, compare with a known good backup or the default template.
-
Validate again Re-run the JSON validation step. Repeat until no errors are reported.
-
Restart Hermes Restart the Hermes agent or reload the configuration (
hermes authmay trigger a reload). -
Verify functionality Try the operation that previously failed (e.g.,
hermes tools,hermes auth check). Success indicates the fix worked.
Pitfalls
- Do not edit protected files like
AGENT.md,MEMORY.md, orSOUL.md; they are managed by the ASI core. - Avoid using non‑JSON tools (e.g.,
catwith special characters) that can corrupt the file further. - If the file appears zero‑length, it may have been truncated; restore from the latest backup or re‑run
hermes authto regenerate. - Some fields (like tokens) are intentionally redacted in logs; do not mistake missing tokens for corruption if they are intentionally omitted.
Verification
After completing the procedure, run:
hermes auth
You should see a successful status (e.g., last_status: "ok" for the active provider) and no JSON parsing errors. Additionally, run a command that uses the configuration (like hermes tools list) to confirm normal operation.
References
- See
references/auth-json-corruption.mdfor a detailed transcript of a real‑world corruption case and the steps taken to resolve it.
04 同仓库更多 Skills
xurl
X/Twitter via xurl CLI: post, search, DM, media, v2 API.
writing-plans
Write implementation plans: bite-sized tasks, paths, code.
whisper
OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.
weights-and-biases
W&B: log ML experiments, sweeps, model registry, dashboards.
watchers
Poll RSS, JSON APIs, and GitHub with watermark dedup.
serving-llms-vllm
vLLM: high-throughput LLM serving, OpenAI API, quantization.