All plugins

error-ledger

Local ledger of failed provider calls (rate limits, overloads, timeouts, context overflows): /errors shows how many attempts failed per model, provider or cause, how many Hermes could retry, and the most common HTTP status, for the last 24 hours, any number of hours or days, the latest session or everything recorded; /errors days puts one day on each line, newest first; /errors recent lists the newest failures (ten, or up to 100 with /errors recent 50) and how long ago each happened. Disclosure — writes failed-attempt metadata (ids, provider, model, HTTP status, Hermes's failure class, exception class name, timings; no error message, request or prompt text) to plugin-data/error-ledger/errors.jsonl, trimmed to 5,000 rows past 2 MB; no network

olric3015Toolsv1.3.0Updated 1 month ago

About

Local ledger of failed provider calls (rate limits, overloads, timeouts, context overflows): /errors shows how many attempts failed per model, provider or cause, how many Hermes could retry, and the most common HTTP status, for the last 24 hours, any number of hours or days, the latest session or everything recorded; /errors days puts one day on each line, newest first; /errors recent lists the newest failures (ten, or up to 100 with /errors recent 50) and how long ago each happened. Disclosure — writes failed-attempt metadata (ids, provider, model, HTTP status, Hermes's failure class, exception class name, timings; no error message, request or prompt text) to plugin-data/error-ledger/errors.jsonl, trimmed to 5,000 rows past 2 MB; no network

error-ledger

A local ledger of the provider calls that failed, and an /errors command that shows which
model or provider they came from and why.

When a provider call fails, Hermes classifies the failure (rate limit, overload, timeout,
context overflow, auth, ...) and then retries, compresses or falls back. Most of the time the
turn still finishes, and the failure is a line that scrolled past. error-ledger subscribes to
api_request_error and keeps one row per failed attempt, so you can see afterwards that one
model was rate-limited forty times today, or that a provider has been answering 529 all morning.

Install

hermes plugins install error-ledger --enable

Requires Hermes 0.21.5 or newer. No API key, no configuration, no dependencies.

Use

/errors             last 24 hours, by model
/errors session     the most recent session that had a failed call
/errors all         everything in the ledger
/errors 7d          the last 7 days (any number of hours or days: 6h, 30d)
/errors providers   add to any of the above to group by provider
/errors reasons     add to any of the above to group by cause
/errors days        add to any of the above to group by day, newest first
/errors recent      add to any of the above to list the newest 10 failures instead
/errors recent 25   the newest 25 instead (1 to 100)
/errors clear       delete the ledger

Example (/errors all, from the end-to-end test on Hermes 0.21.5: one turn that hit a 429 and
one that hit a 500 before recovering):

Provider errors, all recorded:

model       errors  retryable      top reason  top status
fake-model       2          2  rate_limit (1)     429 (1)

One row per failed attempt Hermes reported; the retry after it may have succeeded.
  • errors counts failed attempts, not failed turns: a call Hermes retried three times before
    it gave up is three rows.
  • retryable is how many of them Hermes classified as worth retrying.
  • top reason is Hermes's own classification (rate_limit, overloaded, server_error,
    timeout, context_overflow, auth, billing, ...), with how often it occurred. Grouped
    by reasons, this column shows the model it happened on most instead.
  • top status is the most common HTTP status, when the failure had one.

The tables say how often; /errors recent says when. It lists the newest ten failures of the
chosen window, newest first, so you can tell whether a provider is still failing or stopped an
hour ago:

Provider errors, last 24 hours, newest 3 of 41:

when       model  provider      reason  status  retryable
40s ago  m-large    custom  rate_limit     429        yes
2m ago   m-large    custom  rate_limit     429        yes
3h ago   m-small     other  overloaded     529        yes

/errors days puts one day on each line, newest day first, so a bad day stands out. Days are
your computer's local calendar days, and a day with no failures has no line. Example (the
plugin's own output for a sample week):

Provider errors, last 7 days:

day         errors  retryable       top reason  top status
2026-10-08       5          4   rate_limit (4)     429 (4)
2026-10-07      31         31  overloaded (28)    529 (28)
2026-10-06       3          3   rate_limit (2)     429 (2)
total           39         38  overloaded (28)    529 (28)

One row per failed attempt Hermes reported; the retry after it may have succeeded.

A time window counts back from now (/errors 6h, /errors 7d reasons, /errors 30d days,
/errors 30d recent). /errors recent 50 lists more of them, up to 100.
The ledger keeps its newest 5,000 rows, so a long window on a busy install reaches back only as
far as those rows do.

What it does not see

  • Only failures that reach Hermes's turn loop are reported through api_request_error. Newer
    Hermes reconnects some dropped or failed streams inside its streaming layer without reporting
    them: in the end-to-end test, the recovered 500 is a row on Hermes 0.21.5 and is not on current
    main.
  • Auxiliary calls (titling, compression, vision) use a different hook and are not counted here.

What is recorded

One JSON line per failed attempt in <HERMES_HOME>/plugin-data/error-ledger/errors.jsonl (per
profile):

ts, session_id, platform, provider, model, api_mode, status, reason,
error_type, retryable, retry_count, duration_s.

Not recorded: the error message (only the exception class name), the request, prompts,
responses, base URLs, API keys. The hook hands the plugin the sanitized request and the error
text; the plugin drops both.

The file is bounded: once it passes 2 MB it is rewritten to its newest 5,000 rows.

Security and footprint

  • register() only registers one hook (api_request_error) and one command (/errors).
  • No network access, no subprocesses, no downloads, no credentials read.
  • Writes only errors.jsonl under its own plugin data directory. /errors clear deletes it.
  • Reads no Hermes state other than the hook payload it is handed.
  • The hook is observer-only and never raises: if the ledger cannot be written, the failure is
    logged at debug level and Hermes's own error handling is not affected.
  • Standard library only.

Development

python -m pytest tests                                          # unit tests, no Hermes needed
PYTHONPATH=<hermes checkout> python tests/e2e/error_ledger_e2e.py   # real `hermes chat` turns
hermes plugins validate error-ledger

Changes

  • 1.3.0: /errors recent takes a row count, /errors recent 25 (1 to 100; ten without
    one). What is recorded did not change.
  • 1.2.0: /errors days, one line per day with failures, newest day first; works with
    every window (/errors 30d days). The ledger also reads correctly after a Windows tool has
    saved it with a byte-order mark. What is recorded did not change.
  • 1.1.0: time windows (/errors 6h, /errors 7d, any number of hours or days) and
    /errors recent, the newest ten failures with how long ago each happened. What is recorded
    did not change.
  • 1.0.0: first release.

MIT licensed.

README from the reviewed commit dd64607; it updates when the author re-pins.

More by olric3015

18 plugins by olric3015 in the catalog.

olric3015approval-ledgerLocal ledger of Hermes's approval prompts (dangerous commands, protected writes, MCP consent, smart-mode verdicts): /approval-log shows how often each rule asked, how many prompts were approved, denied or left unanswered, and how long a person took to answer, per rule, per surface or per day, for the last 24 hours, any number of hours or days, the latest session or everything recorded; /approval-log recent lists the newest decisions (ten, or up to 100 with /approval-log recent 50). Disclosure — writes decision metadata (session id, surface, rule label, answer, timings; no command, description or session key) to plugin-data/approval-ledger/approvals.jsonl, trimmed to 5,000 rows past 2 MB; observer only, cannot answer or change an approval; no networkCommunitytoolsolric3015aux-ledgerLocal ledger of Hermes's auxiliary LLM calls (titling, compression, vision, approval): /aux shows tokens, time and failures per task, model, provider or day, or lists the newest calls (ten, or up to 100 with /aux recent 50), for the last 24 hours, any number of hours or days, the latest session or everything recorded. Disclosure — writes auxiliary-call metadata (ids, models, timings, token counts, no prompt or response text) to plugin-data/aux-ledger/calls.jsonl, trimmed to 5,000 rows past 2 MB; no networkCommunitytoolsolric3015b64-ref@b64: context references that decode encoded text for the model: @b64:<value> takes base64 (standard or URL-safe, padded or not), hex (0x48656c6c6f, 48:65:6c, \x48\x65), percent-encoding or a data: URI, told apart by the value's shape or picked with base64:, base64url:, hex: or url:, and attaches what it holds: text as text with control and invisible characters escaped, JSON pretty-printed, gzip and zlib inflated in memory (at most 1 MiB, one layer), and binary named by its format (PNG and GIF with their size, JPEG, PDF, ZIP, ELF, SQLite and more) with a hex dump of its first 128 bytes; notes point out a value that was cut short, padding that is off, a second layer of encoding and a JWT inside; the TUI and Desktop composer preview a value as you type it. Disclosure — reads only the text of the reference itself, using Python's base64, json, urllib.parse and zlib; nothing decoded is run, opened or saved; no network, no files read or written, no subprocess, no credentials read; the plugin keeps no copy of what it decodes, but the @b64: reference and the decoded contents become part of the message sent to your model provider and saved in the session history, so decode only values you are willing to share with it; register() only adds the @b64: reference provider.Communitytoolsolric3015chmod-ref@chmod: context references that explain a file permission mode for the model: @chmod:<mode> takes an octal mode (755, 0755, 4755), ls -l text (-rwsr-xr-x, with the file type letter and a trailing + . or @) or an st_mode as stat and git write it (100644), and attaches both forms, each class's rights in words (directory rights named as such), what setuid, setgid and the sticky bit do and where they are ignored, and notes such as a mode everyone can change, one private to the owner, or a class holding rights the class before it lacks; @chmod:"644 u+x,go-w" applies a symbolic change step by step the way GNU chmod does (u g o a, + - =, r w x X s t, copies like u=g), @chmod:"100644 100755" shows what changes between two modes, and @chmod:"umask 027" shows what new files and directories get; the TUI and Desktop composer preview a mode as you type it. Disclosure — reads only the text of the reference itself, using Python's re; it never looks at a file on disk; no network, no files read or written, no permission changed, no subprocess, no credentials read; the plugin keeps no copy of what you write, but the @chmod: reference and the explanation become part of the message sent to your model provider and saved in the session history; register() only adds the @chmod: reference provider.Communitytools

Developer Mode

Explore built-in Hermes toolsets and actions.

HermesAgent

Developer?

Create a Hermes extension for Desktop or agent workflows.

Developer Section