Tests APIs for Broken Function Level Authorization (BFLA) vulnerabilities where regular users can invoke administrative functions or access privileged API endpoints by directly calling them. The tester identifies admin and privileged endpoints, then attempts to access them with regular user credentials by manipulating…
Use when implementing authorization and access control for FrontMCP tools, resources, prompts, or skills, deciding who may invoke what. Covers the RBAC, ABAC, and ReBAC models and when to choose each; JWT claims mapping per identity provider (Auth0, Keycloak, Okta, Cognito, Frontegg); reusable named authority…
Selects an independent reviewer for OpenFirma design plans and implemented changes without duplicating review work. Use before implementation, before opening or modifying a PR, or when the user requests adversarial review.
Plans substantial OpenFirma changes through proportional repository research, program design, proof obligations, vertical slices, and independent review. Use before implementing behavior, architecture, stable contracts, migrations, trust boundaries, or invariant-owner changes.
Independently reviews an OpenFirma design plan before implementation by inspecting the repository, reconstructing high-risk traces, challenging proof obligations, and reporting preserved findings. Use for Full and Compact planning review.
Manage Slack channel access control — pairing, allowlist, channel opt-in. Use when approving a pairing code, changing the DM policy, editing the user allowlist, or opting a channel in or out. Trigger with "/slack-channel:access", "pair my slack account", "add user to slack allowlist", or "opt in a slack channel".
Author MCP tool-call policy rules without hand-editing access.json. Use when adding, linting, or removing autoapprove/deny/requireapproval rules for the Slack channel's policy engine. Trigger with "/slack-channel:policy", "add a policy rule", "lint my slack policy", or "remove a policy rule".
Create, modify, or manage secrets-dispatcher trust rules. TRIGGER when the user wants to "block", "deny", "allow", "approve", or "ignore" an application or process from accessing secrets, mentions suspicious or unwanted Secret Service requests, asks to add/edit/remove rules in secrets-dispatcher config, or says "add…
Cryptographic receipt signing and Cedar policy enforcement for AI agent tool calls. Every tool invocation produces an Ed25519-signed receipt verifiable offline by anyone, without trusting the issuer.
IDOR (Insecure Direct Object Reference) testing checklist: object ID enumeration, horizontal/vertical privilege escalation, GUID predictability, indirect references via hashes, chained IDOR, and API endpoint IDOR. Use for web app pentests and bug bounty IDOR discovery.
Add a new AccessFlow engine plugin under engines/ / — scaffold the Maven project and QueryEngine SPI implementation, pin the shaded JAR in connectors/ /connector.json, wire DbType + migration + credential gates + i18n + frontend registration + docs/website, and open a PR. Trigger when the user says "add a engine"…
Prepare an AccessFlow release — verify every roadmap item is closed and documented, regenerate website screenshots, mark the milestone released in docs/website/README, then open a chore PR. Refuses to proceed when anything is missing. Trigger when the user says "prepare release vX.Y", "release prep for X.Y", or passes…
Review an existing AccessFlow pull request — check the head out into an isolated worktree, fan out the read-only agents (change summary, conventions, security, frontend, cross-cutting drift), explain what the PR actually does, then merge the findings into one numbered list and let you pick which ones to keep. Never…
Drive the user's real, signed-in browser through Ghostlight's governed tools. Use to open, navigate, inspect, find, click, type, fill forms, screenshot, upload, record, or verify web pages from an agent session, including web-UI testing, rendered-page reading, and form journeys. Prefer these tools over guessing at…
Guides the agent to ask the user for their preferred authentication mode and credentials when they request to connect to the Polaris catalog, rather than using default credentials.