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/john-farina/claude-plugins/conjure-skillgit clone --depth 1 https://github.com/john-farina/claude-pluginsWrote 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/john-farina/claude-plugins/conjure-skill)<a href="https://agentmods.dev/commands/john-farina/claude-plugins/conjure-skill"><img src="https://agentmods.dev/badge/commands/john-farina/claude-plugins/conjure-skill.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 | $0.00073 | $0.03811 |
| Opus 5 | $0.00036 | $0.01906 |
| Sonnet 5 | $0.00015 | $0.00762 |
| Haiku 4.5 | $0.00007 | $0.00381 |
Grade D, and why
conjure-skill scanned grade D with 2 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- Instructions cleared. Describe the new behavior here. --> Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
cat ~/.claude/conjure-config.md 2>/dev/null How it starts
The opening of the file, as written. The whole thing — 384 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/conjure-skill — Create or Update Claude Code Skills
You produce production-quality .md skill files. Every skill you write ships immediately — no follow-up steps required.
Phase 0: Route the Request
Before building anything, verify a skill is the right artifact.
| Question | If YES → use instead |
|---|---|
| Should this fire automatically on an event (file edit, tool call, session start)? | Hook → run /conjure-hook <description> |
| Does Claude spawn this programmatically with isolated context, a different model, or specific tool restrictions? | Agent → run /conjure-agent <description> |
Does it run inline in the main conversation when the user invokes /command? |
Skill → continue ✅ |
Common mislabels:
| Request sounds like… | Actually is… |
|---|---|
| "skill to run lint whenever I edit a file" | Hook — fires automatically |
| "skill to search the codebase with haiku model" | Agent — isolated context, specific model |
| "skill to investigate crashes" | Agent — long-running, spawned programmatically |
| "skill to commit my changes" | Skill ✅ — user-triggered, inline workflow |
| "skill that explains architecture" | Skill ✅ — reference, user loads it |
If it's clearly a hook or agent, stop and say: "This is better as a [hook/agent] — run /conjure-hook or /conjure-agent instead."
Phase 0.5: Load User Preferences
Before doing anything else, read both conjure config files if they exist:
cat ~/.claude/conjure-config.md 2>/dev/null
cat .claude/conjure-config.md 2>/dev/null
Follow all instructions found in ## skills and ## global sections throughout every phase of this command. Project-local instructions (.claude/conjure-config.md) take precedence over global ones when they conflict. If neither file exists, proceed with defaults.
Phase 1: Parse Arguments
Parse $ARGUMENTS:
| Pattern | Mode |
|---|---|
<name> only, file exists |
Update — read + refine existing skill |
<name> only, file missing |
Create — ask one clarifying question then build |
<name> faster |
Speed-optimize existing skill |
<name> simpler |
Simplify existing skill — strip complexity, same core behavior |
<name> detailed |
Expand existing skill — add phases, sub-agents, references |
<name> blank |
Clear body, keep frontmatter, reset instructions |
<name> - <description> |
Create with inline description — build immediately, no question |
| Free-form sentence | Infer name (kebab-case) + description, create immediately |
<name> --git-history |
Git-extract — analyze git history for pattern extraction, then run quality gate before writing final file |
<name> review |
Review — run quality gate on existing skill, report issues, apply fixes |
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 · 384 lines · 73 tokens per session scan D f6ca5953cb25
conjure-skill is a command published in the GitHub repository john-farina/claude-plugins (1 stars, last pushed 3mo ago), licensed MIT. It adds 73 tokens to every session and 3,811 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it D with 2 findings (hidden instructions, reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
sm-sail
Command "sm-sail" from ScienceIsNeato/slop-mop, covering /sm-sail — drive a pr to green, autonomously, the loop, when sail parks on review threads, when to stop before "pr ready" — only two reasons and expect convergence, not one pass.
sm-buff
You usually don't run buff directly — run sm sail. sm sail drives the whole PR to green and calls buff watch / triage for you, stopping only when it needs you to act (see /sm-sail). Reach for sm buff here only for surgical work: inspecting a specific failure, or resolving a single review thread when sail has parked on…
sm-wake-angry-drunk-captain
The last-resort verb. Use it ONLY when the loop is genuinely exhausted: barnacles filed, gates green or truly unfixable, and the single remaining move is a human judgment call no sm verb can make for you.
sm-barnacle
Use when sm itself gives invalid guidance, blocks valid work, produces confusing output, or breaks install/upgrade/refit flow. Do not use this for real target-repo failures; fix those through the normal rail.
sm-init
Run when you find a repo with sm installed but no .sbconfig.json — or after upgrading slopmop to pick up new gates.
sm-refit
Run slop-mop's one-time onboarding remediation rail for this repository.