harnesses

A guide to coding harnesses: command-line tools such as Claude Code, Codex, and Cursor that run a model in an isolated container. The container can plan work, edit files, run commands, install packages, and report back.

In plain words
What is it for?
Use it when creating or configuring a project agent or automation agent, choosing its runtime and model, and checking its sandbox permissions.
Why use it?
It explains where to choose the coding tool, how access is provided, and what the isolated environment can reach. This makes the agent's execution setup easier to understand and control.

Agent

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/tale-project/tale/harnesses
Clone the repo
git clone --depth 1 https://github.com/tale-project/tale
Per session 40 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,180 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 $0.00040 $0.02180
Opus 5 $0.00020 $0.01090
Sonnet 5 $0.00008 $0.00436
Haiku 4.5 $0.00004 $0.00218

Measured yesterday against content hash 78ca3ec28ea6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

harnesses 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 yesterday.

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.

docs/de/platform/agents/harnesses.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.

Ein Harness ist eine mitgelieferte Coding-CLI — Claude Code, Codex, Cursor und weitere —, die dein gewähltes Modell in einem isolierten Container ausführt statt in der gewöhnlichen Chat-Schleife. Das Harness plant, schreibt Dateien, führt Befehle aus, installiert Pakete und berichtet zurück. Im Chat-Composer wählst du kein Harness: Chat wählt nur ein Modell. Das Harness legst du fest, wenn du einen Projekt-Agenten oder einen Automation-Agent-Knoten anlegst — beide Oberflächen nennen das Feld Agent-Laufzeit.

Diese Seite behandelt, welche Harnesses mit Tale kommen, wo du eines bindest, woher der Zugang stammt und was der Container erreichen darf und was nicht. Die Zugänge selbst sind Sache der Organisation — siehe Provider. Unter Einstellungen > Provider zeigt der Tab Harnesses, wie jedes Harness für die Organisation aufgelöst würde.

Wo du ein Harness wählst

Öffne den Tab Agenten eines Projekts und leg einen Agenten an oder bearbeite einen. Der Dialog fragt nach einer Agent-Laufzeit — dem Harness, der Coding-CLI, auf der dieser Agent läuft — neben Modell, Ausrüstung und Anweisungen. Weist du diesem Agenten eine Board-Aufgabe zu, arbeitet er in einer Sandbox auf genau diesem Harness.

In einer Automation trägt ein Agent-Knoten dasselbe Feld Agent-Laufzeit. Erreicht der Workflow diesen Knoten, läuft der Zug auf dem gewählten Harness.

Chat listet keine Harnesses. Die Auswahl im Composer ist nur Modelle; Harness-Arbeit kommt über einen Projekt-Agenten oder einen Automation-Agent-Knoten, nicht über eine Composer-Gruppe.

Was ein Harness-Zug ist

Beschreib die Aufgabe in normaler Sprache — „schreib ein kleines Python-CLI und teste es", „klon dieses Repository und behebe den Fehler aus Issue 42". Die Nachricht geht an das Harness und nicht direkt an das Modell. Das Harness treibt das Modell in einer Schleife im Container an und entscheidet selbst, wann es eine Datei liest, einen Befehl ausführt oder es noch einmal versucht; die Antwort kommt, wenn sein Zug abgeschlossen ist.

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. yesterday First seen · 73 lines · 40 tokens per session scan A 78ca3ec28ea6

Subscribe to this mod's changes

harnesses is an agent published in the GitHub repository tale-project/tale (23 stars, last pushed yesterday), licensed MIT. It adds 40 tokens to every session and 2,180 once invoked, about $0.0002 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 agents, from other repositories

AGENTS

In-depth tutorials on LLMs, RAGs and real-world AI agent applications.

patchy631/ai-engineering-hub · 0 tokens

merge-conflict-resolver

Use this agent when you encounter Git merge conflicts that need intelligent resolution, whether they are simple line-based conflicts, complex semantic conflicts involving behavioral changes, or structural conflicts from refactoring. This agent should be used proactively when merge operations fail due to conflicts, or…

module-federation/core · 264 tokens

bestmode

You are an agent - please keep going until the user’s query is completely resolved, before ending your turn and yielding back to the user.

raphaelmansuy/edgequake · 0 tokens

workflow-debugger

Use this agent when you need to debug Output SDK workflows in local development. Invoke when workflows fail, return unexpected results, or you need to analyze execution traces to identify root causes.

growthxai/output · 40 tokens

electron-e2e-test-runner

Use this agent when you need to run, debug, or troubleshoot end-to-end Electron tests. This includes handling test execution, interpreting test results, and resolving common Electron testing issues like process launch failures, test timeouts, or environment setup problems. Examples:\n\n \nContext: The user is working…

sahithvibudhi/vibe-tree · 365 tokens

harness-interpreter

Reads /tmp/basemind-harden-.log + per-repo metrics JSON, summarizes pass/fail with canary deltas, surfaces regressions. Cheap read-mostly subagent.

Goldziher/basemind · 44 tokens