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/hire-debrief)<a href="https://agentmods.dev/commands/frankxai/starlight-intelligence-system/hire-debrief"><img src="https://agentmods.dev/badge/commands/frankxai/starlight-intelligence-system/hire-debrief/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/hire-debrief"><img src="https://agentmods.dev/badge/commands/frankxai/starlight-intelligence-system/hire-debrief.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 | $0.00090 | $0.03670 |
| Opus 5 | $0.00045 | $0.01835 |
| Sonnet 5 | $0.00018 | $0.00734 |
| Haiku 4.5 | $0.00009 | $0.00367 |
Grade A, and why
hire-debrief 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 10d 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 — 252 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/hire-debrief
This is part of the People Intelligence reference vertical. Composes with Genius Profile + Vision/Brand for company-as-candidate framing.
Load SIP.md, VOICES.md, agents/starlight-hiring.md, skills/people-intelligence/structured-hiring.md, the ICP, the interview architecture, the calibration session record, and (if present) the fit assessment for this candidate. Produce the Debrief Facilitation Script + Decision Record. Hand off to onboarding architecture (if hire-yes) or learning capture (if hire-no).
Disclaimer (non-waivable)
Hiring decisions touch employment law and protected-class considerations. Decision rationale must be rubric-anchored, not feel-anchored, and must not reference protected-class characteristics. This is system architecture, not legal advice. Decision documentation must be retained per jurisdiction-specific requirements; validate with qualified counsel.
This command produces the script + the decision record. The facilitator runs the live session. The decision is rubric-anchored, written down, and retained.
Input
$ARGUMENTS
Flags
--rater-count <3|4|5|6+>— number of raters who interviewed this candidate. Should match calibration session.
Process
-
Disclaim. Open with the non-waivable disclaimer.
-
Locate. Confirm candidate-slug and role-slug. Read the ICP, interview architecture, calibration record, and fit assessment if present.
-
Pre-debrief — collect structured scores. Before the session, every rater submits structured scores via shared doc/tool. Scores are submitted before discussion. Non-negotiable. Loud-voice halo and conformity drift are killed by structure-first.
-
Build the 45-60 minute agenda.
- 0:00 - 0:05 — Frame. Decision rule reminder, bias-pattern primer, structured-scores-before-discussion confirmation.
- 0:05 - 0:15 — Surface dispersion. Where did raters disagree by ≥2 points? Those dimensions get the discussion oxygen.
- 0:15 - 0:35 — Discuss high-dispersion dimensions. Anchor every claim to rubric. Facilitator names bias patterns out loud as they appear.
- 0:35 - 0:45 — Read fit assessment (if present). Supplement to rubric, not replacement.
- 0:45 - 0:55 — Apply decision rule. Hire-or-no-hire per the pre-committed rule. Write the rationale.
- 0:55 - 1:00 — Learning capture. What worked in the loop? What drifted? Feed back into next calibration.
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.
- 10d ago First seen · 252 lines · 90 tokens per session scan A 520482d1f525
hire-debrief is a command published in the GitHub repository frankxai/Starlight-Intelligence-System (8 stars, last pushed today), licensed MIT. It adds 90 tokens to every session and 3,670 once invoked, about $0.0005 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.