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.
npx agentmods add commands/nimbalyst/nimbalyst/social-responsegit clone --depth 1 https://github.com/nimbalyst/nimbalystWrote 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/nimbalyst/nimbalyst/social-response)<a href="https://agentmods.dev/commands/nimbalyst/nimbalyst/social-response"><img src="https://agentmods.dev/badge/commands/nimbalyst/nimbalyst/social-response.svg" alt="Measured on agentmods" 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 | $0.00017 | $0.02070 |
| Opus 5 | $0.00009 | $0.01035 |
| Sonnet 5 | $0.00003 | $0.00414 |
| Haiku 4.5 | $0.00002 | $0.00207 |
Grade A, and why
social-response 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 4d 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 — 167 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/social-response Command
Draft a response to a user reaching out on Discord or another social channel.
User's Message
$ARGUMENTS
What This Command Does
The user pastes a message from Discord, Twitter, Reddit, GitHub Discussions, email, or another social channel. Your job is to:
- Figure out what they're asking for or complaining about
- Determine whether Nimbalyst already supports it
- If yes -- draft a friendly, copy-paste-ready response explaining how
- If no -- decide whether it's a good idea, then ask the user how to proceed using
AskUserQuestion
Process
Step 1: Read the response log
Before drafting anything, read nimbalyst-local/social/social-response.jsonl if it exists. Each line is a previously-approved response. Skim the last 20-30 entries for:
- Tone calibration -- how does the user actually phrase replies?
- Recurring requests -- has someone asked something similar before? What did we tell them?
- Stock answers -- if a feature comes up repeatedly (e.g. "how do I switch themes"), reuse and refine the prior wording.
If the file doesn't exist yet, that's fine -- proceed without it. Do not create the file at this stage; it's only written after the user approves a draft.
Step 2: Understand the request
Read the pasted message. Extract:
- What are they trying to do? (the underlying goal, not just the literal words)
- What channel are they on? (tone shifts: Discord is casual, GitHub is more formal)
- Is this a question, a feature request, a bug report, or general feedback?
If the message is too vague, note that and recommend asking them a clarifying question before responding.
Step 3: Check whether Nimbalyst already supports this
Investigate before drafting. Common moves:
- Read
docs/FEATURE_INVENTORY.md-- the canonical list of what Nimbalyst can do - Search the docs:
docs/,packages/*/CLAUDE.md, and any topic-specific doc listed in the rootCLAUDE.md - Search the codebase with
Grep/Globfor the feature, setting, or behavior they're describing - Check existing trackers with
tracker_list({ search: "..." })-- this may already be a known idea, in-progress work, or a logged decision (including a decision NOT to build it) - Check the changelog (
CHANGELOG.md) -- the feature may have shipped recently
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.
- 4d ago First seen · 167 lines · 17 tokens per session scan A aec0812c71be
social-response is a command published in the GitHub repository nimbalyst/nimbalyst (1,643 stars, last pushed today), licensed MIT. It adds 17 tokens to every session and 2,070 once invoked, about $0.0001 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 2026-08-30.
Other commands, from other repositories
commit
Create a git commit for the current staged (or unstaged) changes using a structured commit message convention.
release-notes
Generate release notes for this project and save them to CHANGELOG.md.
commit-compact
Create a git commit.
review-pr
Review a GitHub PR and post findings.
pre-commit-risk-scoring
Assess pre-commit risk by correlating PagerDuty incidents with current code changes.
git
Git operations with intelligent commit messages and workflow optimization.