Borrowing it
Nothing to install: this file belongs to ManSio/mscodebase-intelligence. 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/ManSio/mscodebase-intelligence/main/.opencode/command/experiment.mdgit clone --depth 1 https://github.com/ManSio/mscodebase-intelligenceWrote 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/commands/mansio/mscodebase-intelligence/experiment)<a href="https://agentmods.dev/commands/mansio/mscodebase-intelligence/experiment"><img src="https://agentmods.dev/badge/commands/mansio/mscodebase-intelligence/experiment/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/commands/mansio/mscodebase-intelligence/experiment"><img src="https://agentmods.dev/badge/commands/mansio/mscodebase-intelligence/experiment.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 | $0.00013 | $0.00610 |
| Opus 5.5 | $0.00005 | $0.00244 |
| Sonnet 5.5 | $0.00003 | $0.00122 |
| Haiku 4.5 | $0.00001 | $0.00061 |
Grade A, and why
experiment 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 26d 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 — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are running the EXPERIMENT phase. A measurement without a control is a story.
Before running
- Write the hypothesis and what would refute it. Not "test if X is faster" — "X is faster than Y when Z, and I will be wrong if the median difference is under 10ms".
- Freeze the input into
experiments/<name>/frozen/with a sha256, BEFORE you look at any result. If you edit the list after seeing the outcome, the result is indistinguishable from hindsight. - Decide the control now. Every experiment needs a case that MUST FAIL. An instrument that has never been seen to fail has not been shown to work.
- Record reproducibility parameters explicitly in the command, never by default: seed, temperature, reasoning budget, model version, variant. A decision made earlier in this session that never made it into the command line did not happen.
While running
- Capture raw output, not a summary. If the run printed 3 lines, keep all 3.
- If the population is empty, the answer is "undeterminable", not 0. An empty input is a different state from a measured zero and must look different on the way out.
- If something fails once, do not retry with the same parameters. Change the hypothesis or the method, and say which.
Afterwards
- Recompute from the raw data. Do not read your own previous verdict back as input. Recompute, then compare.
- Run the gates:
python tools/verification/gates.py --gate population --input '{"population": N, "computed_rate": R}' python tools/verification/run_all.py - Held-out or not? Replaying on the case where the bug was found is
✅ confirmed on known case, never generalization. Mark it exactly. - Count the denominator. "Fixed 3 places" is meaningless. Report
N of M, and how many distinct idioms.
Output
## EXP-<n> — <hypothesis>
**Command:** <exact, copy-pasteable>
**Raw output:** <verbatim, or the path to the frozen file>
**Result:** ...
**Verdict:** CONFIRMED | REFUTED | PARTIAL | HELD-OUT CONFIRMED | UNKNOWN
**Gates:** G1 <verdict> · G2 <verdict> · G4 <verdict>
**Population:** N of M, selected by <rule>
**Negative control:** failed as required — <raw>
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.
- 26d ago First seen · 59 lines · 13 tokens per session scan A 6e2dd0ea6442
experiment is a command published in the GitHub repository ManSio/mscodebase-intelligence (3 stars, last pushed 5d ago), licensed MIT. It adds 13 tokens to every session and 610 once invoked, about $0.0001 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 2026-10-04.
Other commands, from other repositories
index
Index a folder of documents for semantic search.
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.