Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/jketreno/clarenpx agentmods add rules/jketreno/clare/skill-clare2-corpus-captureWrote 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/rules/jketreno/clare/skill-clare2-corpus-capture)<a href="https://agentmods.dev/rules/jketreno/clare/skill-clare2-corpus-capture"><img src="https://agentmods.dev/badge/rules/jketreno/clare/skill-clare2-corpus-capture/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/rules/jketreno/clare/skill-clare2-corpus-capture"><img src="https://agentmods.dev/badge/rules/jketreno/clare/skill-clare2-corpus-capture.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 | $0.00011 | $0.00845 |
| Opus 5 | $0.00005 | $0.00423 |
| Sonnet 5 | $0.00002 | $0.00169 |
| Haiku 4.5 | $0.00001 | $0.00085 |
Grade A, and why
skill-clare2-corpus-capture 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 9d 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 — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.
name: clare2-corpus-capture description: "Capture normalized agent interactions for local CLARE2 distillation" mode: agent
CLARE2 Corpus Capture
Use deterministic lifecycle hooks to send useful agent interactions into the
local CLARE2 corpus. Sessions are captured per-project under a shared user
corpus at ~/.config/clare/corpus, partitioned by project name. Each project
builds its own adapter.
Agent support matrix
| Agent | Capture supported | Notes |
|---|---|---|
| Claude Code | Yes | Hooks in .claude/settings.json |
| Codex | Yes | Hooks in .codex/hooks.json |
| GitHub CoPilot | Unverified | Template written to .github/hooks/clare2-corpus.json; confirm this path matches your gh copilot version |
| Zoo Code | No | VSCode extension API does not expose compatible lifecycle hooks |
Configuration
clare-installer.sh sets CLARE2_CORPUS_ROOT automatically, creates
~/.config/clare/corpus, and symlinks the project's corpus/ directory there
so the pipeline container resolves the same data. Run:
./scripts/clare-installer.sh --update --target /path/to/project
If you need to set it manually:
export CLARE2_CORPUS_ROOT="$HOME/.config/clare/corpus"
CLARE2_SESSION_FILE takes precedence when an existing wrapper or command has
already initialized a session. CLARE2_PROJECT_ID overrides the project name
derived from git rev-parse --show-toplevel basename (used by default).
Text content is redacted and limited to 12,000 characters by default. Set
CLARE2_CAPTURE_MAX_CHARS to a smaller non-negative integer, or 0 to disable
prompt, response, and correction text capture while retaining lifecycle and
tool outcome events.
Reinstall or repair the project hooks with:
./clare/scripts/clare2-install-hooks.sh
Hooks are fail-open. If the corpus is unavailable, normal agent work continues.
Capture Rules
Capture observable evidence:
- submitted user prompts
- final assistant responses when the provider exposes them
- tool names and success/failure outcomes, without raw arguments or output
- session lifecycle events
verify-ci.shresults and corrections already emitted by CLARE
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.
- 9d ago First seen · 115 lines · 11 tokens per session scan A 98c974d886b7
skill-clare2-corpus-capture is a cursor rule published in the GitHub repository jketreno/clare (5 stars, last pushed 1mo ago), licensed MIT. It adds 11 tokens to every session and 845 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 cursor rules, from other repositories
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
cli-error-handling
CLI command error handling patterns.
prefer-assertions-over-defensive-checks
Prefer assertions over defensive checks when data is guaranteed to be valid.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.