pandysp

Get 1-3 independent second opinions from fresh Claude reviewers to strengthen your reasoning before presenting to the user.

Honest self-assessment skill — surfaces incomplete work, missing verifications, bugs, and improvement ideas before shipping.

Black-box verification of any agent output — code, documents, presentations, reports, configs. Examine from the consumer's perspective and show evidence.

Identify and verify unverified claims in the agent's own responses before presenting them as conclusions.

Systematically identify how a plan or design could fail before committing to it.

Generate multiple strong design options with explicit tradeoff profiles, anchored against at least two ideal target architectures.

Before flipping a previously-stated position under pressure, restate its strongest form and the goal it served, then check whether the trigger refutes the core on the same goal — not just a decorative argument or a different axis.

Surface what I think we're doing before acting on it — produces a structured manifest exposing every silent commitment so the user can correct course before any artifact is produced.

Audit error handling for silent failures, inadequate feedback, and inappropriate fallbacks — delegates to the silent-failure-hunter subagent for isolation

Map the terrain before designing — locate, trace, pattern, constrain. Used before any substantive design or planning step. Output is mostly for the agent's grounding; only surprises and hard constraints surface to the user.

Surface and resolve underspecified decisions before designing. Asks targeted questions about scope, edge behavior, trade-offs, integration, constraints, and audience — with defaults — so design choices aren't silently locked around hidden assumptions.

Capture durable lessons from a session before they fade — what surprised you, what friction repeated, what should update CLAUDE.md or notes. Output goes to memory, notes, or global instructions, not just to chat.

Write a durable handoff for work that's about to ship — structured PR descriptions, summaries, memos, or memory notes. Tells the next reader what was built, why, what changed, and what's next, without forcing them to spelunk the diff.

Write the implementation spec — context, approach, files to modify, utilities to reuse, verification — that drives execution after design is chosen.

Write code, comments, tests, or commit messages that read as if the project's own maintainer wrote them, so a contribution is accepted as native rather than rejected as foreign. A foreign-looking patch is a cost to the maintainer even when it's correct.

Route Claude Code's auto-created git worktrees through worktrunk so background and --worktree sessions inherit project hooks (env symlinks, installs, port allocation).

Multi-lens quality audit for docs, code, and other artifacts. 13 quality adjectives as audit lenses with per-domain translation sheets, a workflow-backed finder/verifier pipeline, and lens-yield reporting.