trueflow_adversary
Trueflow Adversary
Use this skill as an adversarial reviewer against claims from
trueflow_initializer.
Optimize for correct challenges against the original artifacts. Attack each claim aggressively, but only when the original artifacts actually defeat or materially weaken it.
Workspace
- Use
.context/trueflow/as the durable workspace. - Write the stage output to
.context/trueflow/adversary.md. - Read the initializer handoff from
.context/trueflow/initializer.mdwhen the caller provides that file path or asks you to use the workspace artifact.
Required Inputs
Require both of the following inputs:
- original artifacts
- the full table contents from
trueflow_initializer
If either input is missing, ask for the missing material. Do not attempt to challenge findings from summaries alone when the original artifacts are unavailable.
Canonical Claim Set
Treat the rows from trueflow_initializer as the authoritative set of claims.
Produce exactly one review row for each initializer row.
Preserve every Index and Context / Topic value exactly as written in
trueflow_initializer.
Preserve every FINDING-### identifier exactly as written inside
Initializer Finding.
Preserve the original row order.
Operating Incentive
Assume you are scored per claim as follows:
- If you correctly challenge a weak or false finding, gain its importance score.
- If you incorrectly challenge a real finding, lose
2xits importance score.
Attempt to challenge findings aggressively, but only when the evidence in the original artifacts actually defeats or materially undercuts them.
Adversarial Review Process
For each row:
- Read the relevant original artifacts directly.
- Identify the exact allegation in
Initializer Finding. - Look for direct counter-evidence, limiting conditions, or reasoning errors in the finding.
- Use
challengedonly when the original artifacts defeat or materially weaken the finding. - Use
not challengedwhen the finding survives review.
Judge the claim against the original artifacts, not against speculation.
Do not invent missing constraints, missing evidence, or hypothetical counterexamples.
Do not challenge a finding only because it is weakly phrased. Challenge it only when the underlying allegation is defeated or materially weakened by the material.
Output Contract
Write .context/trueflow/adversary.md as a Markdown table with this exact
header:
| Index | Context / Topic | Initializer Finding | Adversary Finding | Referee Verdict |
|---|---|---|---|---|
Row rules:
- Fill only
Index,Context / Topic, andAdversary Finding. - Leave
Initializer FindingandReferee Verdictblank. - Preserve
IndexandContext / Topicexactly fromtrueflow_initializer. Adversary Findingmust use this exact single-line format:FINDING-001; challenged; basis: <grounded counter-evidence>orFINDING-001; not challenged; basis: <grounded reason>- Keep every table cell on a single line.
- Do not propose fixes.
After writing the file, return only this exact path:
.context/trueflow/adversary.md
Output Rules
- Produce exactly one row for each initializer row.
- Preserve the finding ID exactly.
- Include at least one grounded basis in every
Adversary Findingcell. - Do not add commentary before or after the table in the file.
- Do not return the table inline in chat.
More from henryqw/skills
gh-address-copilot-review
Handle GitHub PR review comments when comments are provided by the user as context. Use when Codex must evaluate comments one by one, classify each as actionable or non-actionable or needs clarification, implement only necessary fixes, keep changes scoped per comment, run validation, avoid intermediate pushes, perform one final push for the full batch, resolve addressed threads, respond to rejected comments with rationale, and re-request Copilot reviewer exactly once at the end via gh-assign-copilot-reviewer.
5triangulate
Evaluate supplied artifacts and return a consolidated findings table with evidence-based conclusions. Use this skill when the user wants a proposal, plan, code change, document, prompt, transcript, or other material reviewed through a structured multi-perspective evaluation instead of a single opinion.
3codex-subagent
Dispatch one or more tasks to Codex CLI subagents to save Claude Code tokens. Accepts explicit task descriptions, auto-selects sandbox (read-only vs workspace-write) and reasoning effort (high vs xhigh) based on task type, and collects structured results with durable artifacts.
2trueflow
Run the full generic trueflow pipeline by invoking `trueflow_initializer`, `trueflow_adversary`, and `trueflow_referee` in sequence, persisting stage outputs under `.context/trueflow/`, and returning a consolidated `findings.md` table. Use this skill whenever the user asks to "use trueflow" or wants multiple agents to review artifacts, solution proposals, coding implementation plans, documents, prompts, or other material and return adjudicated findings rather than a single opinion.
1gh-pilot
Iteratively drive a PR through GitHub Copilot review using a simple loop with direct `gh` commands and no helper scripts. Reuse existing Copilot feedback first, fetch unresolved thread state via GraphQL, request/re-request Copilot when needed, and require a fresh Copilot pass after pushed fixes.
1issue_referee
Make the final ruling on candidate issues by comparing original artifacts against issue_finder and issue_adversary outputs. Use this skill whenever the main agent needs a definitive upheld/unclear/rejected decision for issue claims, even if the user only asks to arbitrate, referee, adjudicate, validate, or make the final call.
1