adversarial-implementation

adversarial-implementation is a skill for Claude Code from xdg/xdg-claude. It costs 52 tokens per session (1,682 once invoked), scanned A, original, Apache-2.0.

A workflow for implementing TODO.md tasks in small units with separate implementation, testing, review, and human-verification steps.

In plain words
What is it for?
Use it to work through TODO items one at a time, run lint and tests, obtain adversarial reviews, handle tasks needing human checks, and commit completed subsections.
Why use it?
It reduces the risk of accepting incomplete or incorrect work by checking each checkbox through independent verification before marking its section done.

Skill for Claude Code

Written for Claude Code: disable-model-invocation in frontmatter. Also seen: mentions CLAUDE.md; mentions subagents.

Part of the adversarial-implementation plugin — 1 skill shipped together

Good fit Use it to work through TODO items one at a time, run lint and tests, obtain adversarial reviews, handle tasks needing human checks, and commit completed subsections.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/xdg/xdg-claude/adversarial-implementation
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.

Any agent
npx skills add xdg/xdg-claude --skill adversarial-implementation
Clone the repo
git clone --depth 1 https://github.com/xdg/xdg-claude

Made for: Claude Code.

Or install adversarial-implementation, the plugin that ships this one along with the rest of its 1 skill.

Wrote 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.

agentmods badge for adversarial-implementation

README.md
[![agentmods](https://agentmods.dev/badge/skills/xdg/xdg-claude/adversarial-implementation.svg)](https://agentmods.dev/skills/xdg/xdg-claude/adversarial-implementation)
Your own site
<a href="https://agentmods.dev/skills/xdg/xdg-claude/adversarial-implementation"><img src="https://agentmods.dev/badge/skills/xdg/xdg-claude/adversarial-implementation.svg" alt="Measured on agentmods" height="20"></a>
Per session 52 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,682 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 5.1 $0.00052 $0.01682
Opus 5 $0.00026 $0.00841
Sonnet 5 $0.00010 $0.00336
Haiku 4.5 $0.00005 $0.00168

Measured 8d ago against content hash 15f034e06cc6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

adversarial-implementation 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 8d 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.

adversarial-implementation/skills/adversarial-implementation/SKILL.md · 174 lines

How it starts

The opening of the file, as written. The whole thing — 174 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Task: $ARGUMENTS (If no task is provided, read TODO.md, if it exists. Implement the next incomplete subsection. If there is no TODO.md file, prompt the user for instructions.)

Adversarial Implementation Protocol

When implementing from TODO.md:

  • Implementation unit: Individual checkboxes (keeps subagent sessions manageable)
  • Completion unit (aka "subsection"): The immediate parent heading of a group of checkboxes—i.e., the most specific heading that directly contains them, not any ancestor.

Example TODO structure:

### 3.1 OAuth Setup
#### 3.1.1 Configure provider
- [ ] Add client ID
- [ ] Add client secret
#### 3.1.2 Implement callback
- [ ] Handle redirect

Here, 3.1.1 and 3.1.2 are completion units (each is a subsection directly containing checkboxes). 3.1 OAuth Setup is NOT a completion unit—it's a parent grouping.

A completion unit is complete only when ALL its checkboxes are checked—including any that require human verification.

Isolated subagents

Whenever this guide says to invoke an isolated subagent, fall back to a general agent if the isolated subagent doesn't exist. In either case, use model 'opus' and effort 'low' for subagents unless otherwise instructed by the user.

Phase 1: Plan the Subsection

For the next incomplete Phase subsection:

  1. Read ALL checkboxes within the subsection
  2. Categorize each checkbox:
    • Automatable: Code changes, automated tests
    • Human-required: Browser interactions, visual verification, manual testing
  3. Process automatable items one at a time (or in small logical batches) through Phases 2-4 before moving to the next item
  4. Track which items are complete within the subsection
  5. If subsection has NO automatable items, skip directly to Phase 6

Phase 2: Implement Single Item via Subagent

For each automatable checkbox:

  1. Spawn an isolated task subagent (if one exists, or a general subagent otherwise) with this prompt structure:
   Task: [specific checkbox item]
   Context files: TODO.md, any other relevant specs or instructions (OMIT code
   files; the subagent can find these on its own)
   Constraints: [project rules, style guide refs]
   Acceptance criteria: [the specific checkbox, explicitly stated]
   Output: Implement and report a concise summary of what you changed.
  1. Do NOT provide the full codebase—give minimal necessary context so you minimize use of the main context. Include: TODO.md, relevant specs, file paths to modify. Exclude: file contents the subagent can read itself, unrelated modules.

Read the full file on GitHub · 174 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. 8d ago First seen · 174 lines · 52 tokens per session scan A 15f034e06cc6

Subscribe to this mod's changes

adversarial-implementation is a skill published in the GitHub repository xdg/xdg-claude (20 stars, last pushed today), licensed Apache-2.0. It adds 52 tokens to every session and 1,682 once invoked, about $0.0003 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 2026-08-30.

Related

Other skills, from other repositories

research-engineer

An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.

davila7/claude-code-templates · 43 tokens

tika-eval-compare

Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".

apache/tika · 50 tokens

neuron-evaluation-engineer

Create and run AI evaluations with datasets, assertions, and output drivers in Neuron AI. Use this skill whenever the user mentions evaluation, testing AI systems, creating evaluators, dataset-driven testing, assertion-based validation, or wants to measure AI system performance. Also trigger for tasks involving…

neuron-core/neuron-ai · 77 tokens

jetson-validate-image

Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.

NVIDIA/skills · 50 tokens

atmos-validation

Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.

cloudposse/atmos · 31 tokens

skill-benchmark

Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.

HoangNguyen0403/agent-skills-standard · 16 tokens