What is adaptation-parity?

A workflow for connecting an AI agent to BenchFlow, an evaluation system, through its agent communication protocol, then checking that it behaves the same inside tests and when run on its own.

Use it when adding an agent to the repository or checking one already there: implement the required protocol server and registration, map its streamed events, and verify evaluation-versus-standalone behavior.

It reduces the risk that an agent passes evaluations but behaves differently in real use because the two environments invoke it differently.

adaptation-parity — Skill for Claude Code and Codex by benchflow-ai

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 benchflow-ai/agents --skill adaptation-parity

Clone the repo

git clone --depth 1 https://github.com/benchflow-ai/agents

Made for: Claude Code, Codex.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/benchflow-ai/agents/adaptation-parity/github.svg)](https://agentmods.dev/skills/benchflow-ai/agents/adaptation-parity)

Your own site

<a href="https://agentmods.dev/skills/benchflow-ai/agents/adaptation-parity"><img src="https://agentmods.dev/badge/skills/benchflow-ai/agents/adaptation-parity/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/benchflow-ai/agents/adaptation-parity"><img src="https://agentmods.dev/badge/skills/benchflow-ai/agents/adaptation-parity.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 58 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 987 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.

Model	Per session	Once invoked
Fable 5.1	$0.00058	$0.00987
Opus 5.5	$0.00023	$0.00395
Sonnet 5.5	$0.00012	$0.00197
Haiku 4.5	$0.00006	$0.00099

Measured 2d ago against content hash e4c57008a6a1, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

adaptation-parity 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.

The scan reads SKILL.md. This mod also ships 11 executable files (scripts/acp_capture.mjs, scripts/acp_driver.mjs, scripts/acp_fixture.mjs, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

skills/adaptation-parity/SKILL.md · 77 lines

How it starts

The opening of the file, as written. The whole thing — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Adaptation + Parity

This repo's thesis: the agent you ship is the agent you benchmark. This skill makes that real in two steps — adapt an agent to BenchFlow, then verify parity (it behaves the same inside the eval harness as standalone).

1. Adapt (agent → BenchFlow ACP)

An agent becomes a BenchFlow agent by speaking ACP over stdio and registering via the public register_agent extension point. Pattern (see ai-sdk/acp, ai-sdk/harness-pi for worked examples):

server.mjs — a pure-JS ACP-over-stdio server. Newline-delimited JSON-RPC; stdout = protocol only (logs → stderr). Handle initialize, session/new, session/set_model, session/prompt (run the agent loop, stream session/update events, return {stopReason, usage}), session/cancel. Map the agent's stream → ACP: text→agent_message_chunk, reasoning→agent_thought_chunk, tool call→tool_call (name+args in title), tool result→tool_call_update. Route the model at OPENAI_BASE_URL (the gateway).
register.py — register_agent(name, install_cmd, launch_cmd, protocol="acp", api_protocol="openai-completions", env_mapping={BENCHFLOW_PROVIDER_*→agent vars}, acp_model_format="bare", requires_env=[]). install_cmd bootstraps node, base64-deploys server.mjs, and npm-installs deps in the sandbox.

For prod parity, the same server.mjs/loop must be the production runtime — no eval-only reimplementation. That's the whole point.

Scaffold a new adapter: python scripts/scaffold_ai_sdk_agent.py <name> (prints a ready-to-edit package skeleton mirroring ai-sdk/acp).

2. Verify parity (inside == outside)

Behavioral parity = given the same model responses, the agent sends the same request and takes the same actions whether driven inside BenchFlow or standalone. Verify at two levels:

(a) Wire parity — drive the agent's ACP server against a capturing mock upstream, twice: standalone, and through BenchFlow's gateway. Diff the upstream requests.

Read the full file on GitHub · 77 lines

Files

What ships with it

13 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.

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.

2d ago Changed · +6 lines e4c57008a6a1
1mo ago First seen · 71 lines · 58 tokens per session scan A 95c7be829572

Subscribe to this mod's changes

adaptation-parity is a skill published in the GitHub repository benchflow-ai/agents (11 stars, last pushed 2d ago), licensed Apache-2.0. It adds 58 tokens to every session and 987 once invoked, about $0.0002 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 2026-08-31.

Building agents acp ai-agents coding-agent llm-evaluation opencode swe-bench

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openinterpreter/openinterpreter · 114 tokens

plugin-creator

Create and scaffold plugin directories for Codex with a required .codex-plugin/plugin.json, optional plugin folders/files, valid manifest defaults, and personal-marketplace entries by default. Use when Codex needs to create a new personal plugin, add optional plugin structure, generate or update marketplace entries…

openinterpreter/openinterpreter · 86 tokens

codex-pr-body

Update the title and body of one or more pull requests.

openinterpreter/openinterpreter · 17 tokens

remote-tests

Testing against remote executors in integration tests.

openinterpreter/openinterpreter · 12 tokens

update-v8-version

Update Codex's pinned v8 / rustyv8 versions, validate the release-candidate path, and investigate failed V8 canary or artifact builds. Use when asked to bump V8, update rustyv8 artifacts, prepare or validate a V8 release candidate, check v8-canary, or diagnose why a V8 version update no longer builds.