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/bborbe/semantic-search/explorergit clone --depth 1 https://github.com/bborbe/semantic-searchWhat 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 | $0.00021 | $0.00767 |
| Opus 5 | $0.00010 | $0.00383 |
| Sonnet 5 | $0.00004 | $0.00153 |
| Haiku 4.5 | $0.00002 | $0.00077 |
Grade A, and why
explorer 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 — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Usage
/semantic-search:explorer obsidian workflow
/semantic-search:explorer "kafka backup strategy" --max-iters=12
/semantic-search:explorer --server=personal raw schema cdb pipeline
/semantic-search:explorer claude code agents --workspace=~/explorer-runs/claude
When to use
/semantic-search:search— one-shot ranked list, no traversal/semantic-search:research— multi-source synthesis from top-N hits, no link traversal, no external exploration/semantic-search:explorer— goal-directed exploration with self-assessed stopping: planner first understands what the user is actually asking and decomposes it into concrete sub-questions, generator explores (vault pages, semantic searches, external repos/URLs), evaluator decides when each sub-question is answered
Process
1. Parse arguments
- First non-flag string →
<topic>(required; show usage and STOP if missing) --server=<label>→ scope semantic search to one MCP server (personal,work,(default)); omit to query all--max-iters=<N>→ hard cap on generator iterations (default20)--workspace=<dir>→ explicit workspace path; default$HOME/.semantic-search-explorer/<topic-slug>-<unix-ts>/
2. Compute the workspace path
Default: ephemeral — agent creates via mktemp -d -t semantic-search-explorer (under $TMPDIR; OS cleans up). The synthesis is embedded in the caller report so nothing is lost when the workspace vanishes.
Pass --workspace=<dir> to preserve the workspace for inspection (debug "why did saturation fire on iter 4", inspect intermediate notes/, etc.).
3. Invoke the explorer-assistant agent
Task tool with:
subagent_type: 'semantic-search:explorer-assistant'
prompt: 'Explore topic <topic>. Workspace: <WORKSPACE>. Server scope: <server or "all">. Max iterations: <N>.'
The agent runs the Planner → Generator → Evaluator loop, writes synthesis to <WORKSPACE>/synthesis.md, and returns a short report.
4. Present the result
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 · 58 lines · 21 tokens per session scan A 41a3c47adcc7
explorer is a command published in the GitHub repository bborbe/semantic-search (2 stars, last pushed 6d ago), licensed BSD-2-Clause. It adds 21 tokens to every session and 767 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 2026-08-31.
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.
constitution
Create or update the project constitution from interactive or provided principle inputs.
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.