All plugins

interim-ledger

Local ledger of the messages models write in the middle of a turn, before their final answer (the text that comes with a tool call, which on chat platforms can arrive as messages of their own): /interim-log shows how many each model wrote, in how many turns, how many per such turn, their median length and how long ago the last one was, per model, provider, surface or day, for the last 24 hours, any number of hours or days, the latest session or everything recorded; /interim-log recent lists the newest messages (ten, or up to 100 with /interim-log recent 50). Counted even where display.interim_assistant_messages hides them. Disclosure — writes message metadata (model, provider, surface, step of the turn, length in characters, 12-digit hashes of the session and turn ids; never the message text) to plugin-data/interim-ledger/interim.jsonl, trimmed to 5,000 rows past 2 MB; observer only, cannot delay, change or hide a message; no network

olric3015Toolsv1.0.0Updated 1 month ago

About

Local ledger of the messages models write in the middle of a turn, before their final answer (the text that comes with a tool call, which on chat platforms can arrive as messages of their own): /interim-log shows how many each model wrote, in how many turns, how many per such turn, their median length and how long ago the last one was, per model, provider, surface or day, for the last 24 hours, any number of hours or days, the latest session or everything recorded; /interim-log recent lists the newest messages (ten, or up to 100 with /interim-log recent 50). Counted even where display.interim_assistant_messages hides them. Disclosure — writes message metadata (model, provider, surface, step of the turn, length in characters, 12-digit hashes of the session and turn ids; never the message text) to plugin-data/interim-ledger/interim.jsonl, trimmed to 5,000 rows past 2 MB; observer only, cannot delay, change or hide a message; no network

interim-ledger

A local ledger of the messages models write in the middle of a turn, before their final answer,
and an /interim-log command that shows which models write them, how many per turn and how
long they are.

Many models say something before they call a tool: "Let me check the config first", "Now I'll
run the tests". Hermes calls these mid-turn (interim) messages. In the terminal they scroll past
with the tool output, but on Telegram, Discord or Slack each one can arrive as a message of its
own, so a chatty model fills the chat with them. interim-ledger subscribes to
on_interim_message, which Hermes fires for each of them, and keeps one row per message. You can
see afterwards that one model writes three of them per turn and another almost none, and decide
whether to switch models or turn them off with display.interim_assistant_messages.

Install

hermes plugins install interim-ledger --enable

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

Use

/interim-log             last 24 hours, by model
/interim-log session     the most recent session
/interim-log all         everything in the ledger
/interim-log 7d          the last 7 days (any number of hours or days: 6h, 30d)
/interim-log providers   add to any of the above to group by provider (or surfaces, days)
/interim-log recent      add to any of the above to list the newest 10 messages instead
/interim-log recent 25   the newest 25 instead (1 to 100)
/interim-log clear       delete the ledger

Example (/interim-log all, from the end-to-end test: two hermes chat turns, the first with one
message before its tool call, the second with two):

Mid-turn messages, all recorded:

model       messages  turns  per turn  median chars    last
fake-model         3      2       1.5            57  3s ago

Counted even where display.interim_assistant_messages hides them.
  • messages counts mid-turn messages, and turns the turns that had at least one.
  • per turn is messages divided by turns: how many a model writes when it writes any.
  • median chars is the middle length of those messages, in characters.
  • last is how long ago the newest one was. The model with the most messages comes first.

Hermes fires the hook whether or not your chat shows these messages, so a model's habit is
counted even where display.interim_assistant_messages is off.

/interim-log recent lists the newest ten messages of the chosen window, newest first
(/interim-log recent 25 lists 25, up to 100). From the same run:

Mid-turn messages, all recorded, newest 3 of 3:

when         model  surface  step  chars  streamed
3s ago  fake-model      cli     2     52        no
3s ago  fake-model      cli     1     57        no
7s ago  fake-model      cli     1     57        no

surface is where the turn ran (cli, telegram, discord, ...), step which model call
of the turn wrote it, and streamed whether Hermes reports the text as already streamed live.

/interim-log days puts one day on each line, newest day first; days are your computer's local
calendar days. A time window counts back from now (/interim-log 6h, /interim-log 7d days).
session and all take precedence over a time window. 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

  • A turn whose model calls tools without writing anything first: there is no message to count.
  • The final answer of a turn: only messages before it are mid-turn messages.

What is recorded

One JSON line per mid-turn message in
<HERMES_HOME>/plugin-data/interim-ledger/interim.jsonl (per profile):

ts, session, turn, surface, model, provider, iteration, chars, streamed.

The text of the message is never written, only its length. session and turn are the first 12
hex digits of a SHA-256 of Hermes's session and turn ids, enough to tell them apart.

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 (on_interim_message) and one command (/interim-log).
  • Observer only: Hermes runs the hook on its own background thread, so it cannot delay, change
    or hide a message. Each hook call appends one line to the ledger, and nothing else.
  • No network access, no subprocesses, no downloads, no credentials read.
  • Writes only interim.jsonl under its own plugin data directory. /interim-log clear deletes
    it.
  • The hook never raises: if the ledger cannot be written, the failure is logged at debug level.
  • Standard library only.

Development

python -m pytest tests                                               # unit tests, no Hermes needed
PYTHONPATH=<hermes checkout> python tests/e2e/interim_ledger_e2e.py   # real hermes chat turns
hermes plugins validate interim-ledger

Changes

  • 1.0.0: first release.

MIT licensed.

README from the reviewed commit ed94f43; 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