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 skills/rahmanef63/si-coder-agent/use-si-codernpx skills add rahmanef63/si-coder-agent --skill use-si-codergit clone --depth 1 https://github.com/rahmanef63/si-coder-agentWhat 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 | $0.00090 | $0.03384 |
| Opus 5 | $0.00045 | $0.01692 |
| Sonnet 5 | $0.00018 | $0.00677 |
| Haiku 4.5 | $0.00009 | $0.00338 |
Grade A, and why
use-si-coder scanned grade A with 1 finding 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 2d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
Full `@convex-dev/auth` setup — generating `JWT_PRIVATE_KEY` + `JWKS`, the required backend env-var table, the PBKDF2 password-hashing override (Scrypt/bcrypt time out on the Dokploy proxy), the `ConvexClientProvider` "r How it starts
The opening of the file, as written. The whole thing — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SI Coder Auto Deploy (legacy /use-si-coder)
This skill automates the entire lifecycle of creating a GitHub repository and deploying full-stack apps to a Dokploy server via the single monolithic scripts/deploy.js. It is the original one-shot pipeline; the modular /sc-* skills (/sc-all, /sc-dokploy, /sc-convex, …) now cover the same ground in surgical pieces, but this monolith remains supported.
CORE MANDATES FOR THE AI (LEARNED LESSONS)
To guarantee Zero Human Involvement (the user just wants to receive the final working URL):
- Self-Hosted Convex by Default: NEVER use Clerk unless explicitly asked. ALWAYS use
@convex-dev/auth. ALWAYS include adocker-compose.ymlfor self-hosting Convex alongside the frontend. - Build Safety: Do NOT run
npx convex codegeninside theDockerfile. You MUST generate the types locally (npx convex dev --once) and commit theconvex/_generatedfolder to Git before deploying. - No Prompts / Dependency Hell: Always use
npm install --yes --legacy-peer-deps. If a scaffolded template is too complex (bloated), wipe it and start fresh withnpx create-next-appto avoid endless TypeScript errors. - Exact Cloning: If asked to clone a website, you MUST replicate its actual layout and structure, not just build a generic admin dashboard. Fetch the website to understand its structure.
- Dokploy Idempotency: The deployment script handles existing apps, composes, and domains safely. Do NOT delete or recreate existing domains if Dokploy rejects them (it means they are already configured).
- Convex Admin Key Sync: When generating or rotating a self-hosted Convex admin key, immediately update the Dokploy Compose environment and the repo's local env file so they stay in sync. Use the project's actual env name for that template or app, and do not invent a second secret name.
- Clerk MCP for Clerk Apps: If a target project already uses Clerk, preserve it, do not swap it to
@convex-dev/auth, and use Clerk MCP (clerkathttps://mcp.clerk.com/mcp) for SDK snippets and integration patterns.
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.
- 2d ago First seen · 152 lines · 90 tokens per session scan A 0d24d76b975c
use-si-coder is a skill published in the GitHub repository rahmanef63/si-coder-agent (14 stars, last pushed 5d ago), licensed MIT. It adds 90 tokens to every session and 3,384 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
convex-self-hosting
Integrate Convex static self hosting into existing apps using the latest upstream instructions from get-convex/self-hosting every time. Use when setting up upload APIs, HTTP routes, deployment scripts, migration from external hosting, or troubleshooting static deploy issues across React, Vite, Next.js, and other…
setup-convex-worktrees
Set up, repair, or explain isolated Convex development deployments for Git worktrees and coding agents, including per-worktree env bindings, temporary deployment creation, baseline data seeding, frontend port isolation, and AGENTS.md guardrails. Use when separate checkouts are clobbering one shared Convex dev…
garage
Operate Joel's Garage S3-compatible object storage on three-body. Use for Garage, Convex Garage, Garage S3, three-body Garage, port 39100, Garage buckets, and MinIO-to-Garage migration.
deployhq
Deploy code, manage servers, and automate infrastructure via the DeployHQ CLI (dhq). Use when the user wants to deploy, check deployment status, manage projects/servers, or interact with the DeployHQ platform.
Convex Cron Jobs
Scheduled function patterns for background tasks including interval scheduling, cron expressions, job monitoring, retry strategies, and best practices for long-running tasks.
kagent-dev
Development guide for kagent's v1alpha3 Harness and AgentTemplate CRDs, AgentInstance gRPC control plane, upstream A2A integration, Substrate runtime provisioning, tests, generation, and PR workflow. Use for any implementation, debugging, review, or CI task in the kagent repository.