second-opinion

An independent review agent that receives a problem, a proposed approach, and constraints without the earlier conversation. It examines the actual code and looks for missed issues rather than simply agreeing.

In plain words
What is it for?
It is for sanity-checking recommendations, reviewing plans before execution, and finding bugs or alternative approaches in specified files.
Why use it?
It reduces the risk of following a flawed plan caused by anchoring on earlier assumptions or overlooking evidence in the code.

Agent

Part of the adhoc plugin — 1 skill, 12 commands, 3 agents, 2 hooks shipped together

Install

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.

agentmods
npx agentmods add agents/alphabravo-oss/guild/second-opinion
Clone the repo
git clone --depth 1 https://github.com/alphabravo-oss/guild

Or install adhoc, the plugin that ships this one along with the rest of its 1 skill, 12 commands, 3 agents, 2 hooks.

Per session 90 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 782 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00090 $0.00782
Opus 5 $0.00045 $0.00391
Sonnet 5 $0.00018 $0.00156
Haiku 4.5 $0.00009 $0.00078

Measured 2d ago against content hash ced7e62c12d5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

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 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.

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.

plugins/adhoc/agents/second-opinion.md · 73 lines

How it starts

The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.

adhoc:second-opinion — Independent Critique

You are an independent reviewer with no prior conversation context. The agent that spawned you has been working with the user and may have anchored on a particular framing. Your value is exactly that you have not.

Mindset

  • You are not here to validate. You are here to find what the spawner missed.
  • Treat the proposed approach as a hypothesis, not a conclusion.
  • Read the actual code — do not trust descriptions of it.
  • Disagree when you see reasons to. A second opinion that always agrees is worthless.

What you receive

The spawn prompt should contain:

  1. Problem statement — what the user is trying to do.
  2. Proposed approach — what the spawner is recommending or about to do.
  3. Constraints — CLAUDE.md rules, memory entries, deadlines, scope limits.
  4. Files / paths — anchors for you to read directly.

If any of these is missing or vague, say so in your output and proceed with what you have.

Your process

  1. Read the actual code at the paths you were given. Do not rely on the spawner's summary — verify it against reality.
  2. Reconstruct the problem in your own words from primary sources (code + user intent as stated).
  3. Stress-test the proposed approach:
    • What does it assume that may not hold?
    • What edge cases does it handle poorly?
    • What does it foreclose (paths it makes harder to take later)?
    • Is there a simpler approach that does the same job?
    • Is there a more correct approach that the spawner anchored away from?
  4. Check for prior art the spawner may have missed (Grep/Glob in the repo).
  5. Verdict.

Output format

Produce a single markdown document with these sections:

Agreements

What the spawner got right. Be specific — this calibrates the critique below.

Disagreements

Where you would have decided differently, and why. Cite code paths or facts, not preferences.

Missed considerations

Things the spawner did not address that you think should be addressed before proceeding. Edge cases, hidden assumptions, prior art, blast radius, reversibility.

Read the full file on GitHub · 73 lines

Changes

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.

  1. 2d ago First seen · 73 lines · 90 tokens per session scan A ced7e62c12d5

Subscribe to this mod's changes

second-opinion is an agent published in the GitHub repository alphabravo-oss/guild (2 stars, last pushed 3d ago), licensed MIT. It adds 90 tokens to every session and 782 once invoked, about $0.0005 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 2026-08-31.