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/frankxai/Starlight-Intelligence-SystemWrote 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/frankxai/starlight-intelligence-system/distill-insights)<a href="https://agentmods.dev/commands/frankxai/starlight-intelligence-system/distill-insights"><img src="https://agentmods.dev/badge/commands/frankxai/starlight-intelligence-system/distill-insights/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/frankxai/starlight-intelligence-system/distill-insights"><img src="https://agentmods.dev/badge/commands/frankxai/starlight-intelligence-system/distill-insights.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 | $0.00049 | $0.01677 |
| Opus 5 | $0.00024 | $0.00839 |
| Sonnet 5 | $0.00010 | $0.00335 |
| Haiku 4.5 | $0.00005 | $0.00168 |
Grade A, and why
distill-insights 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/distill-insights
Load SIP.md, VOICES.md, agents/starlight-secondbrain.md, agents/starlight-genius.md, skills/memory/insight-distillation/SKILL.md, and genius/profile-<slug>.md if it exists. Distill the person's captures from the specified window into named frameworks using the ≥3-occurrence threshold. Propose Genius Profile evolutions where warranted.
Input
$ARGUMENTS
Required args: --person <name> --window <week|month|quarter>. Parse slug from name (kebab-case). Parse window as one of week (last 7 days), month (last 30 days), quarter (last 90 days).
When this command fires
- Weekly distillation — the lightweight pass inside the weekly review (or standalone)
- Monthly distillation — the dedicated 90-minute session; the primary distillation cadence
- Quarterly integration — the half-day session that crosses distillates with the Genius Profile
When this command does NOT fire
- Daily capture — use
/capture-daily - Full weekly review orchestration — use
/orchestrate-brain - One-time knowledge reorganization — use
/reclaim-knowledge - Excavating a Genius Profile from scratch — use
/discover-genius
Process
-
Load captures from window. Read all files matching
second-brain/captures/<slug>/*.mdwhere the date falls in the window. Also readsecond-brain/inbox/(inbox items have no date constraint — they were unrouted captures waiting). -
Cluster by theme. Read every capture in full (not skim). Group captures that touch the same underlying theme. Name clusters in the person's own voice — not "client-handling patterns" but whatever phrase the captures themselves suggest. Clusters can overlap; one capture can sit in two clusters.
-
Count occurrences per cluster. Apply threshold:
- ≥3 occurrences → eligible for framework elevation (step 4)
- 2 occurrences → hold for next window; note in cluster file
- 1 occurrence → keep as anecdote or drop
-
Draft frameworks for eligible clusters. For each ≥3 cluster, write a framework draft using the structure from
skills/memory/insight-distillation/SKILL.md:
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.
- 11d ago First seen · 150 lines · 49 tokens per session scan A e06700e38208
distill-insights is a command published in the GitHub repository frankxai/Starlight-Intelligence-System (8 stars, last pushed today), licensed MIT. It adds 49 tokens to every session and 1,677 once invoked, about $0.0002 per session on Opus 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
memories
View and manage learned memories.
mpm-session-resume
Load context from paused session.
forget
Delete specific memories.
learn
Add new learning to memory.
gbu-retro
Post-session retrospective — harvest this session's lessons into durable doctrine.
consolidate
Write a compact checkpoint summary of the current frontier.