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 skills add JairoTorregrosa/jaiskills --skill second-opiniongit clone --depth 1 https://github.com/JairoTorregrosa/jaiskillsWrote 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/skills/jairotorregrosa/jaiskills/second-opinion)<a href="https://agentmods.dev/skills/jairotorregrosa/jaiskills/second-opinion"><img src="https://agentmods.dev/badge/skills/jairotorregrosa/jaiskills/second-opinion/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/skills/jairotorregrosa/jaiskills/second-opinion"><img src="https://agentmods.dev/badge/skills/jairotorregrosa/jaiskills/second-opinion.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 | $0.00223 | $0.03407 |
| Opus 5.5 | $0.00089 | $0.01363 |
| Sonnet 5.5 | $0.00045 | $0.00681 |
| Haiku 4.5 | $0.00022 | $0.00341 |
Grade A, and why
second-opinion 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 19d 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 — 217 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Second opinion
Request: $ARGUMENTS
Mode by first word: review → review mode · adversarial (or challenge) → adversarial mode ·
judge → judge mode · anything else (or advise) → advise mode. --provider claude anywhere in
the request → skip Codex, go straight to the fallback path and label the result
"same-provider (requested)".
Codex advises, reviews and judges; it never implements. Every call is read-only whatever the
user's Codex config says (it may set danger-full-access). Never ask Codex for a patch to apply,
never pass it a writable sandbox. Claude verifies what it reports and does any fixing.
Paths below are relative to this skill's directory, ${CLAUDE_SKILL_DIR} (other harnesses: the
absolute path of the folder holding this SKILL.md). Shell state does not survive between Bash
calls: set REPO in the same call that runs the script.
Transport (both modes)
Always scripts/ask_codex.sh (the Codex MCP server no longer exists: removed in codex-cli
0.154). Never hand-rolled codex flags. The script runs codex exec --sandbox read-only --skip-git-repo-check --ephemeral -C "$REPO" -o <out>/last-message.md - (or codex exec review
with sandbox_mode="read-only"; review ignores --output-schema, openai/codex#38545), disables Codex memories so no unrelated session leaks in, ignores execpolicy .rules, lifecycle
hooks and the notify command (each can run outside the sandbox), pins review_model for native
review, checks
install and codex login status first, and enforces a timeout.
Model and effort. Default model gpt-6-astra when Codex's catalog
(~/.codex/models_cache.json) lists it, else the config default; SECOND_OPINION_MODEL or -m
overrides. Default effort high (SECOND_OPINION_EFFORT or -e; the user's config may default
to low). Use xhigh for adversarial reviews of risky changes; max/ultra only when asked.
REPO=$(git rev-parse --show-toplevel 2>/dev/null || pwd) && \
"${CLAUDE_SKILL_DIR}/scripts/ask_codex.sh" -C "$REPO" [--schema FILE] [-e high|xhigh] \
[--keep-session] [--resume SESSION_ID|last] [--review uncommitted|base:REF|commit:SHA] \
[-t 540] [-m MODEL] < prompt.md # --review: < /dev/null (it takes no instructions)
What ships with it
8 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 19d ago Changed · +42 lines · -6 tokens per session fee79809427b
- 20d ago First seen · 175 lines · 229 tokens per session scan A adca603ec559
second-opinion is a skill published in the GitHub repository JairoTorregrosa/jaiskills (5 stars, last pushed 14d ago), licensed MIT. It adds 223 tokens to every session and 3,407 once invoked, about $0.0009 per session on Opus 5.5. 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 2026-09-27.
Other skills, from other repositories
review
Review code, specifications, or designs for quality, bugs, security, and best practices. Use when the user wants feedback on recent work, a pull request, or any artifact.
re0-merge
Review and land an external contribution the way this suite does: gate it against the thesis, land it with the author's credit intact, complete a new skill rather than merging it raw, then approve, credit, and explain before closing. Use when reviewing a pull request, as any collaborator or maintainer, not only the…
sip
After you create or change an artifact or skill, taste-test it with our own skills instead of trusting your in-session judgment — recursive self-improvement, made automatic. Use right after writing or editing anything, before calling it done, committing, or handing it off.
prism
Split one artifact — a claim, plan, or file — across 2 to 5 independent lenses, one per genuinely distinct failure mode (correctness, security, readability, cost, adversarial-user), and return their convergence: where they agree, where they disagree, and the single next question that resolves the disagreement. Use…
hate
Attack a plan, design, or argument like you want it to fail before you commit real effort — return the single load-bearing objection and the cheapest experiment that would prove it matters, not a checklist. User-invoked on purpose: a hate-it reflex always in the agent's reach would bias it toward demolition. Weighs…
fabric-spec
Starts a persistent Pi Fabric spec supervisor that audits the main session against a feature design spec and steers only when a requirement lacks verified evidence. Use for strict, unblocked spec compliance while the main agent keeps full freedom to orchestrate.