Borrowing it
Nothing to install: this file belongs to parcadei/ContinuousClaudeV4.7. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/parcadei/ContinuousClaudeV4.7/main/.claude/skills/upgrade-harness/SKILL.mdgit clone --depth 1 https://github.com/parcadei/ContinuousClaudeV4.7Wrote 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.
[](https://agentmods.dev/skills/parcadei/continuousclaudev4.7/upgrade-harness)<a href="https://agentmods.dev/skills/parcadei/continuousclaudev4.7/upgrade-harness"><img src="https://agentmods.dev/badge/skills/parcadei/continuousclaudev4.7/upgrade-harness.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00022 | $0.00896 |
| Opus 5 | $0.00011 | $0.00448 |
| Sonnet 5 | $0.00004 | $0.00179 |
| Haiku 4.5 | $0.00002 | $0.00090 |
Grade A, and why
upgrade-harness 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.
What it actually says
Walk through adding a new external function to the ouros sandbox harness. External functions are the sandbox's only way to interact with the outside world — each one pauses Python execution, runs the real operation on the host, and returns the result.
Use when adding capabilities like "perplexity search to the sandbox", "Slack messages from sandbox", "database query function", or "extend sandbox with X".
The harness lives at tools/ouros_harness.py with extensive built-in functions: exa_search, nia_search variants, llm_call, agent_call, read_file, write_file, glob_files, run_command, research_package, pipeline_start, pipeline_done, plus security infrastructure for async web APIs, filesystem operations, and shell commands.
Start by understanding the request. Ask the user what function they want, what parameters it takes, whether it calls external APIs (async) or runs locally (sync), and what security constraints it needs (path restrictions, API keys, rate limits).
Read the current harness to understand existing patterns. Bridge functions are defined after imports, before SECURITY_POLICY. The SECURITY_POLICY dict defines allowed paths/commands. EXTERNAL_FUNCTIONS dict maps names to handlers.
Write the bridge function following established patterns. For async functions that call web APIs, create an async function plus sync wrapper:
async def _call_new_function(param1, param2="default"):
"""Bridge to the real API."""
script_dir = Path(__file__).parent
sys.path.insert(0, str(script_dir))
from my_api_module import api_call
if not param1:
return {"error": "param1 is required"}
return await api_call(param1=param1, param2=param2)
def _call_new_function_sync(*args, **kwargs):
"""Sync wrapper -- ouros external functions must be sync."""
return asyncio.run(_call_new_function(*args, **kwargs))
For sync functions that run locally:
def _call_new_function(param1, param2="default"):
"""Description of what this does."""
if not _check_path_allowed(param1, SECURITY_POLICY["new_allow"]):
return {"error": f"new_function denied: '{param1}' is outside allowed directories"}
try:
result = some_operation(param1, param2)
return result
except Exception as e:
return {"error": f"new_function failed: {e}"}
Bridge functions must return JSON-serializable dicts or strings, catch all exceptions returning {"error": "..."}, keep parameters simple (strings, ints, bools, lists), and provide sync wrappers for async functions since ouros calls sync only.
Add security policy if the function accesses filesystem, network, or shell. Add rules to SECURITY_POLICY dict then use _check_path_allowed() or write custom checks. Security principles: deny by default, allowlists over denylists, check before executing, fail closed on errors.
Register in EXTERNAL_FUNCTIONS dict where the key name becomes what sandbox code calls:
EXTERNAL_FUNCTIONS = {
# ... existing functions ...
"new_function": _call_new_function_sync, # or _call_new_function for sync
}
Test end-to-end with smoke tests (does it work?), security tests (blocks unauthorized access?), and error handling tests (fails gracefully?). Run from project directory:
echo 'result = new_function("test_arg"); print(result)' | python tools/ouros_harness.py
Add test cases to test_ouros_harness.py if it exists. Update the function table in .claude/CLAUDE.md with the new capability. Install any Python package dependencies into /tmp/ouros/.venv/bin/pip install package-name and document what was installed. Copy any separate scripts to the harness directory.
Present completion checklist: bridge function written, security policy added, registered in EXTERNAL_FUNCTIONS, smoke tested, security tested, CLAUDE.md updated, dependencies installed and documented.
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.
- 8d ago First seen · 74 lines · 22 tokens per session scan A 5519c711032a
upgrade-harness is a skill published in the GitHub repository parcadei/ContinuousClaudeV4.7 (48 stars, last pushed 4mo ago), licensed MIT. It adds 22 tokens to every session and 896 once invoked, about $0.0001 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.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…
chronicle
Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…