Borrowing it
Nothing to install: this file belongs to okuzpe/midas-harness. 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/okuzpe/midas-harness/main/.agents/skills/midas-sandbox/SKILL.mdgit clone --depth 1 https://github.com/okuzpe/midas-harnessWrote 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/okuzpe/midas-harness/midas-sandbox)<a href="https://agentmods.dev/skills/okuzpe/midas-harness/midas-sandbox"><img src="https://agentmods.dev/badge/skills/okuzpe/midas-harness/midas-sandbox/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.
<a href="https://agentmods.dev/skills/okuzpe/midas-harness/midas-sandbox"><img src="https://agentmods.dev/badge/skills/okuzpe/midas-harness/midas-sandbox.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00093 | $0.02621 |
| Opus 5 | $0.00046 | $0.01311 |
| Sonnet 5 | $0.00019 | $0.00524 |
| Haiku 4.5 | $0.00009 | $0.00262 |
Grade A, and why
midas-sandbox 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 160 lines — stays where its author put it; the contents beside it link to each section on GitHub.
midas-sandbox — real skill dry-run lab (engine only)
Guard + state:
<paths.engine>/templates/skill-state-ritual.md(+AGENTS.md§ Safety / Path resolution). Engine-only. Isolation + cost:sandbox/README.md. Runner:node scripts/sandbox-run.mjs. Tally:<paths.engine>/templates/audit-checklists.md§ Parseable tally lines.
Dispatches a real Midas skill (unmodified, from harness/skills/) against
sandbox/example-product/, on a Task subagent pinned to composer-2.5 (never
composer-2.5-fast). Findings are proposals; nothing under harness/skills/ or
harness/rules/ is edited here.
Does / Does not
| Does | Does not |
|---|---|
| Abort unless cwd is midas-harness engine | Run on product installs / template trees |
Always reset then require env exit 0 before any Task |
Trust a dirty example-product/ or engine harness/state.yaml |
| Run the target skill in the composer-2.5 Task | Spawn nested Task / midas-builder on another model |
Log [SANDBOX AUTO-DECISION] for every substituted AskQuestion |
Auto-apply findings to harness source |
Write sandbox/findings/<date>-<mode>.md + MIDAS_SANDBOX_RESULT: |
Advance real engine stage / gates |
Run sandbox-run grade (disk oracles) after the Task |
Let composer grade its own homework |
Feedback loop (touch → detect → classify → re-smoke)
When the diff touches harness/skills/** or harness/rules/**, the default lab is --skill /
--smoke, not --all. Findings stay proposals (no writes under harness/skills / harness/rules
until a later turn after human OK). After a harness-gap patch, re-run --smoke until grade
passes. Skip --smoke only when the fixture cannot exercise the change (docs-only prose,
engine-only skill itself). Adapters / VERSION → /midas-align.
Engine guard (hard)
- Confirm engine:
package.json→"name": "midas-harness"andscripts/test.mjsexists. - Otherwise print: ABORT —
/midas-sandboxis engine-only; use/midas-verify→ STOP.
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.
- 5d ago Changed · +23 lines fda99853096a
- 10d ago First seen · 137 lines · 93 tokens per session scan A 0a5e9d2931a0
midas-sandbox is a skill published in the GitHub repository okuzpe/midas-harness (2 stars, last pushed 6d ago), licensed Apache-2.0. It adds 93 tokens to every session and 2,621 once invoked, about $0.0005 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 2026-08-31.
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.
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".
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…
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.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.
skill-benchmark
Benchmark AI skill effectiveness by measuring implementation quality against legacy constraints.