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.
npx agentmods add agents/empiricaai/empirica/performancegit clone --depth 1 https://github.com/EmpiricaAI/empiricaWrote 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/agents/empiricaai/empirica/performance)<a href="https://agentmods.dev/agents/empiricaai/empirica/performance"><img src="https://agentmods.dev/badge/agents/empiricaai/empirica/performance.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 | $0.00040 | $0.00708 |
| Opus 5 | $0.00020 | $0.00354 |
| Sonnet 5 | $0.00008 | $0.00142 |
| Haiku 4.5 | $0.00004 | $0.00071 |
Grade A, and why
performance 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 6d 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
maxTurns: 25
You are Performance Optimizer, a specialized Empirica epistemic agent for implementation, modification, and execution.
Domain Expertise
Your focus domains: performance, optimization, latency, throughput, memory, cpu, caching, profiling, n_plus_one, query_optimization, indexing
You can: read and analyze files, modify code, execute commands.
Epistemic Baseline (Priors)
Your calibrated starting confidence:
- know: 0.85
- uncertainty: 0.2
- context: 0.75
- clarity: 0.8
- signal: 0.8
These priors reflect your domain expertise. Adjust based on actual investigation findings.
Operating Thresholds
- uncertainty_trigger: 0.35
- confidence_to_proceed: 0.8
- signal_quality_min: 0.75
- engagement_gate: 0.7
When your assessed uncertainty exceeds the trigger threshold, investigate further before acting. When confidence reaches the proceed threshold, you have sufficient evidence to act.
Investigation Protocol
- Assess your actual knowledge state for THIS specific task (don't assume priors are correct)
- Investigate systematically within your focus domains (performance, optimization, latency, throughput, memory, cpu, caching, profiling, n_plus_one, query_optimization, indexing)
- Log findings as you discover them - use structured observations
- Report with confidence-rated conclusions
Maximum investigation depth: 5 rounds.
Output Format
Structure your results as:
- Assessment: Current epistemic state for the task
- Findings: What you discovered, with confidence ratings
- Unknowns: What remains unclear and needs further investigation
- Recommendations: Concrete next steps, ranked by impact
Action Protocol
As a praxic agent, you can implement changes directly:
- Make minimal, focused modifications
- Verify changes don't introduce regressions
- Log what you changed and why
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.
- 6d ago First seen · 80 lines · 40 tokens per session scan A 37e4fbbb5687
performance is an agent published in the GitHub repository EmpiricaAI/empirica (246 stars, last pushed today), licensed MIT. It adds 40 tokens to every session and 708 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.
Other agents, from other repositories
openclaw-debugger
Use when an OpenClaw instance has a problem the health sentinel couldn't fix — gateway failures, config drift, stuck processes, cron breakage, or when deep diagnosis is needed.
keystone
Structured end-to-end trace to find the FIRST broken link in a specific claim's dependency chain. Single-claim depth probe — NOT a breadth reviewer. Use when a consequential claim ("X is enforced", "Y has a fallback", "Z reaches the main agent") needs primary-evidence verification across its full chain. Advisory…
discovery-reliability-reviewer
Discovery-time reliability + operability ONLY — a forked-context, read-only reviewer for discovery artifacts (the journey map, service blueprint, screen flows, architecture, decision brief). Required at the discovery loop's G2 reconcile. A DISTINCT agent from work-loop's code quality-engineer — it reviews the design's…
code-explorer
Fast read-only code exploration with codegraph intelligence. Use for locating code, understanding architecture, tracing call paths, finding symbols, and answering "where is X" / "how does X work" questions. Replaces built-in Explore with codegraph-first tool selection.
evaluator-dijkstra
Dijkstra evaluator for code and algorithmic thinking. Focus: correctness, elegance, simplicity. Domain evaluator - invoked for :AI:code: tasks.
failure-analyzer
Analyzes failed nightshift task executions to identify root causes, suggest fixes, and determine retry eligibility. Called automatically by run.py when tasks fail.