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 commands/ecro/embedeval/researchgit clone --depth 1 https://github.com/Ecro/embedevalWhat 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 | $0.00000 | $0.00669 |
| Opus 5 | $0.00000 | $0.00334 |
| Sonnet 5 | $0.00000 | $0.00134 |
| Haiku 4.5 | $0.00000 | $0.00067 |
Grade A, and why
research 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 yesterday.
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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/research — Research & best practices
Gather information, explore options, recommend an approach. Step 1 of the
embedeval 5-stage workflow. Informs /myplan — does not create any file.
Communication
- Direct, evidence-based, no preamble.
- When reasoning is flawed, say so with specific evidence.
- Don't fold on pushback without new evidence.
- Lead with concerns before agreement.
Usage
/research <topic or task>
Use Opus with ultrathink for thorough analysis.
Context
- Repo root:
/home/noel/embedeval - Published docs:
docs/*.md(METHODOLOGY, BENCHMARK-, LLM-EMBEDDED-) - Existing plans:
plans/PLAN-*.md - Prior corrections:
CLAUDE.md→ "Learned Corrections" - Domain references:
docs/LLM-EMBEDDED-FAILURE-FACTORS.md— 42 factors, 6 categoriesdocs/LLM-EMBEDDED-DEVELOPMENT-GUIDE.md— end-to-end workflowdocs/LLM-EMBEDDED-CONSIDERATIONS.md— production failure patterns
Steps
-
Codebase scan — use
Agent(subagent_type="Explore")for broad exploration, or Grep/Glob directly for narrower searches. Inspectsrc/embedeval/,cases/, relevant tests. -
Prior embedeval work — grep
plans/PLAN-*.mdand CLAUDE.md corrections for related topics. Read every hit that looks relevant. -
External research — WebSearch for best practices, Context7 MCP for official library docs. Cite sources.
-
Synthesize — identify 2–3 approaches with pros/cons, complexity, risk, estimated effort.
Output (screen only)
## Research: <topic>
### Problem
<what, why, constraints, success criteria>
### Approaches
#### Option A (recommended): <name>
- Pros / Cons / Complexity / Risk / Effort
#### Option B: <name>
- Pros / Cons / Complexity / Risk / Effort
### Prior embedeval work
<PLAN-*.md hits or CLAUDE.md corrections worth reusing/avoiding>
### External references
<web / Context7 / docs with links>
### Pitfalls
1. <pitfall> — <how to avoid>
### Recommendation
Proceed with Option <X>. Rationale: <2–3 sentences>.
### Next step
`/myplan <suggested-task-slug>`
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.
- yesterday First seen · 90 lines · 0 tokens per session scan A e1d506e087f7
research is a command published in the GitHub repository Ecro/embedeval (11 stars, last pushed 26d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 669 tokens. 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 commands, from other repositories
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.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
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.