Showing results for "allowlist"
Design enforcement: lockfile required, checksum verification, artifact signing, and dependency allowlists/denylists. Include alerting on suspicious updates.
Author: Assistant
Category: safe-self-improving-ai | Model: gpt-5.2
Design an agent that tunes security scanners (SAST rules, allowlists) based on confirmed findings and false positives. Require approvals for any rule weakening.
Author: Assistant
Category: safe-self-improving-ai | Model: gpt-5.2
Design sanitization so retrieved web/docs cannot inject instructions. Include instruction isolation, allowlisted tool use, and red-team tests for prompt injection.
Author: Assistant
Category: safe-self-improving-ai | Model: gpt-5.2
Design containment by limiting action scope: allowlisted domains, read-only modes, rate limits, and staged privileges. Include how to measure whether containment is effective.
Author: Assistant
Category: recursive-ai-safety | Model: GPT-5.2
Design domain packs: specialized source allowlists, stricter citation rules, and domain-specific extraction (e.g., trials, filings). Include how to swap packs safely.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Create a crawl plan: seed selection, depth limits, domain allowlists, URL patterns, and prioritization heuristics. Include a method to estimate crawl cost and coverage.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Design defenses against prompt injection for tool-using agents: content provenance, allowlists, tool policy, and sandboxing. Include a suite of adversarial prompts for regression testing.
Author: Assistant
Category: recursive-ai-safety | Model: GPT-5.2
If browser automation is required, design a policy: strict allowlists, minimal interaction, rate limits, and recording of steps. Include risk assessment and a user-facing disclaimer template.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Create a safe-defaults checklist for the tool layer: deny-by-default, explicit allowlists, safe parameter validation, output filtering, and timeouts. Include common failure modes.
Author: Assistant
Category: recursive-ai-safety | Model: GPT-5.2
Design layered guardrails: policy rules, tool allowlists, output filters, and human escalation. Include regression tests to prevent guardrail drift during iterations.
Author: Assistant
Category: agent-architecture | Model: GPT-5.2
Draft defenses for tool-using agents: content sanitization, domain allowlists, URL reputation, and read-only sandboxes. Provide red-team prompts and pass/fail gates.
Author: Assistant
Category: safety-security | Model: gpt-4o
Build a fast secrets scanner for files and git diffs (regex + entropy). Flags: --path, --staged, --baseline, --allowlist, --exit-on-find. Print remediation tips and rotate guidance.
Author: Assistant
Category: cli-tool | Model: gpt-4o