Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/ashlrai/ashlr-pluginWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/ashlrai/ashlr-plugin/ashlr-compact)<a href="https://agentmods.dev/commands/ashlrai/ashlr-plugin/ashlr-compact"><img src="https://agentmods.dev/badge/commands/ashlrai/ashlr-plugin/ashlr-compact/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/ashlrai/ashlr-plugin/ashlr-compact"><img src="https://agentmods.dev/badge/commands/ashlrai/ashlr-plugin/ashlr-compact.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00021 | $0.00676 |
| Opus 5.5 | $0.00008 | $0.00270 |
| Sonnet 5.5 | $0.00004 | $0.00135 |
| Haiku 4.5 | $0.00002 | $0.00068 |
Grade A, and why
ashlr-compact scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Read ~/.ashlr/session-history/<sessionId>.jsonl using the ashlr__read tool (or Bash if the path resolves) and then run the following analysis logic:
-
Find the session ID: It is in
~/.ashlr/last-project.jsonunder thesessionIdkey. Read that file first. -
Read the session history: Read
~/.ashlr/session-history/<sessionId>.jsonl. Each line is a JSON object with fields:ts,tool,sizeBytes,contentSha8,turn,sessionId. -
Compute current turn count: Count the total number of lines in the history file — this is
currentTurn. -
Classify entries by staleness:
turnDelta = currentTurn - entry.turn- Fresh:
turnDelta < 5(freshness 1.0) - Stale:
5 ≤ turnDelta < 15(freshness 0.5) - Very stale:
turnDelta ≥ 15(freshness 0.2)
-
Aggregate:
- Total stale results: count of entries where
turnDelta ≥ 5 - Total stale bytes: sum of
sizeBytesfor stale entries - Per-tool breakdown: group by
tool, show count + bytes
- Total stale results: count of entries where
-
Print the report to the user in this format:
ashlr-compact: Stale result analysis ───────────────────────────────────── Session turns: <currentTurn> Stale results: <count> (turnDelta ≥ 5) Stale volume: ~<KB> KB By tool: <tool> <count> results · ~<KB> KB ... Recompression opportunity: ~<staleKB> KB could be freed by re-running these reads now (the fresh results would replace stale ones in context). Suggested actions:Then list the top 5 stale entries (by sizeBytes descending) and suggest re-running the specific tool that produced them (e.g.,
re-run ashlr__read on <implied subject>if tool is ashlr__read or Read). -
Inject recompression hint as your next assistant message context:
"ashlr: The above stale results are from 5+ turns ago. If any are still needed, consider re-running the relevant reads now with fresh context. Stale results from earlier turns remain in the conversation history and count against the context window even when no longer load-bearing."
If the history file does not exist or is empty, tell the user: "No ashlr session history found — the tracker starts recording after the first ashlr__read or ashlr__grep call."
If total stale bytes < 5 KB, tell the user: "Session looks clean — less than 5 KB of stale tool output. No recompression needed yet."
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 55 lines · 21 tokens per session scan A b63dc5db681a
ashlr-compact is a command published in the GitHub repository ashlrai/ashlr-plugin (3 stars, last pushed 2d ago), licensed MIT. It adds 21 tokens to every session and 676 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
recall-save
Generate / overwrite .recall/context.md with Recall's local offline summarizer.
show
Show this project's saved Recall context (.recall/context.md).
ccc-context-budget
Show how full the context window is, which components use the most of it, and when to save the session and compact. Use when the session feels slow, long or near its context limit.
promote
Show what the tape is holding for your word, and promote the ones you approve into canon. Agent inference never becomes canon on its own — this is the door, and your hand is on it.
bifrost-candidates
Review memory candidates recorded during sessions, and promote the good ones into shared team memory.
pin
Break checkpoint — status display, session log, optional MEMORY.md update.