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/josephfung/trimkit/prod-debugnpx skills add josephfung/trimkit --skill prod-debuggit clone --depth 1 https://github.com/josephfung/trimkitWhat 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 | $0.00048 | $0.01428 |
| Opus 5 | $0.00024 | $0.00714 |
| Sonnet 5 | $0.00010 | $0.00286 |
| Haiku 4.5 | $0.00005 | $0.00143 |
Grade A, and why
prod-debug 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.
- Health check: curl https://<domain>/api/health How it starts
The opening of the file, as written. The whole thing — 179 lines — stays where its author put it; the contents beside it link to each section on GitHub.
prod-debug
Invocation
Two modes:
/prod-debug— load context and enter debugging mode/prod-debug bootstrap— (re)build schema.md and containers.md from source files
On /prod-debug
Step 1: Find project root
Determine the project root — use the current working directory, or if inside a git repo, use git rev-parse --show-toplevel.
Step 2: Check for data directory
Check whether .claude/prod-debug/config.json exists at the project root.
- If missing: tell the user that prod-debug is not set up for this project yet, and suggest running
/prod-debug bootstrapafter creating.claude/prod-debug/config.json. - If present: continue to Step 3.
Step 3: Load context files
Read all three data files using the Read tool:
{project_root}/.claude/prod-debug/schema.md{project_root}/.claude/prod-debug/containers.md{project_root}/.claude/prod-debug/prod-env.md
If any file is missing, note it and proceed with what's available.
Step 4: Announce
Report a brief summary:
Prod-debug loaded.
Schema: {N} tables
Containers: {M} services
Env: {hostname or domain from prod-env.md}
Ready — ask me to query the DB, inspect logs, or diagnose an issue.
Step 5: Enter debugging mode
Use the loaded context to assist with:
DB queries — Build SQL using the pre-loaded schema. No \d or INFORMATION_SCHEMA queries needed — the schema is already in context. For queries on remote prod, prefix with the SSH + docker exec command from prod-env.md.
Container operations — Use container names and commands from containers.md. Common patterns:
- View logs:
docker logs {service} --tail 100 -f - Shell in:
docker exec -it {service} sh(or bash) - Restart:
docker restart {service}
SSH to prod — Use the SSH alias/command from prod-env.md. Remote docker commands are typically run via ssh {alias} docker ...
Stale schema — If the user mentions a table or column that isn't in schema.md, flag it and suggest running /prod-debug bootstrap to rebuild from migrations.
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 · 179 lines · 48 tokens per session scan A b10a89aef1a3
prod-debug is a skill published in the GitHub repository josephfung/trimkit (3 stars, last pushed 3mo ago), licensed MIT. It adds 48 tokens to every session and 1,428 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
dynamic-workflow-patterns
Pattern taxonomy, agent role combinations, model routing, unit-of-work sizing, review-findings settling, and resilience discipline for Claude Code dynamic workflows. ALWAYS load this skill before authoring or running any Workflow tool script, and ALWAYS load it when the user mentions "workflow" or "ultracode" in any…
serena-tool-selection
MANDATORY tool selection protocol for Serena LSP tools vs Claude Code built-in tools. ALWAYS use when your tools list includes any mcpserena tools. This skill OVERRIDES default tool selection behavior for code navigation tasks.
a11y-audit
../../../engineering-team/a11y-audit/skills/a11y-audit/SKILL.md.
agent-memory
../../../engineering/agent-memory/skills/agent-memory/SKILL.md.
agile-product-owner
../../../product-team/agile-product-owner/skills/agile-product-owner/SKILL.md.
app-store-optimization
../../../marketing-skill/skills/app-store-optimization/SKILL.md.