# tc_harness

> tc_harness — zapsyq-tc-harness. Use this tool when you need to stabilize interactions with small language models, connecting them to various tools while handling issues like long context compression, malformed JSON repair, and automatic retry of broken turns. It solves problems related to tool integration, context management, and error handling, providing a reliable interface for inputs like tool commands and outputs like processed results. Ideal for use cases involving multiple tool calls, error-prone interactions, or complex workflows.

Canonical page: https://skillsregistry.net/skills/zapsyq-tc-harness  
JSON: https://api.skillsregistry.net/v1/skills/zapsyq-tc-harness

## Description

TC_Harness —— 让小模型稳定调用工具的智能体框架 / Tool-calling agent harness for small language models — connect MCP tools, compress long context, auto-repair malformed JSON/tool history, and retry broken turns automatically.

## Trust

- **Trust score (0–1):** 0.50
- **Verification tier:** unverified

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **License:** Apache-2.0
- **Updated:** 2026-10-03

## Source

- **Source listing:** [GitHub](https://github.com/zapsyq/tc_harness)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "zapsyq-tc-harness"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/zapsyq-tc-harness` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/zapsyq-tc-harness/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt