MILTONADINA
@MILTONADINA
Public Skills
goal-loop
by MILTONADINA
Execute the spec-tests-implement-verify cycle for one atomic task at a time. Use after plan-decomposition has produced a task list. This is the workhorse loop - the agent implements, runs tests, reflects on failure, retries with bounded iteration. Halts mechanically when budget, iteration count, or scratchpad stasis ceilings hit. See constitution/LOOP.md for full protocol.
unsigned-missing-sig-fixture
by MILTONADINA
Test fixture (F.09) — SKILL.md without an accompanying .sig file. Default install rejects with exit 2; --allow-unsigned --rationale="..." proceeds.
unsigned-mismatched-sig-fixture
by MILTONADINA
Test fixture (F.09) — SKILL.md whose accompanying .sig signs different content. NFR-F3 hash check catches this even with --allow-unsigned.
karpathy-guidelines
by MILTONADINA
Behavioral guidelines that reduce common LLM coding mistakes. Apply whenever writing, modifying, or reviewing code. Don't assume requirements, don't overcomplicate, don't refactor unrelated code, and define success criteria before implementing. Use this skill on every non-trivial coding task to avoid bloat, silent assumptions, and completion-bias.
prompt-injection-defense
by MILTONADINA
Sanitize all external content before it enters the agent's context window. Use whenever the agent ingests documents, web pages, RAG results, tool outputs, file contents, emails, or any data from outside the user's direct typed input. Defense against OWASP ASI01 (Goal Hijacking) and ASI04 (Indirect Prompt Injection) - the #1 and #4 risks for 2026. Triggers on any file read, fetch, search, or external API response.
karpathy-guidelines
by MILTONADINA
Behavioral guidelines that reduce common LLM coding mistakes. Apply whenever writing, modifying, or reviewing code. Don't assume requirements, don't overcomplicate, don't refactor unrelated code, and define success criteria before implementing. Use this skill on every non-trivial coding task to avoid bloat, silent assumptions, and completion-bias.
supabase-rpc-functions
by MILTONADINA
Supabase RPC function security patterns -- SECURITY DEFINER with SET search_path discipline, REVOKE EXECUTE FROM public for least-privilege exposure, JSON-arg shape drift between client and server contracts. Triggers on Supabase RPC function authoring (CREATE OR REPLACE FUNCTION, rpc( client calls). OWASP ASI02 (Tool Misuse) defense -- RPCs that bypass RLS are a privileged-backdoor path.
rust-error-handling
by MILTONADINA
Rust error-handling patterns and anti-patterns -- unwrap() in request hot paths, panic! instead of returning a Result, leaking error internals via Debug derive on error types containing privileged state. Patterns covered include thiserror for typed errors, anyhow for application-layer error propagation, and error-payload sanitization at API boundaries. Triggers on Rust error-prone code authoring (unwrap, expect, panic!, Result<, > patterns). General security hygiene at the language-runtime layer.
supabase-rls-policies
by MILTONADINA
Supabase Row-Level Security policy patterns and failure modes -- policy bypass via service-role key in client code, USING-vs-WITH-CHECK confusion, missing INSERT policies that leave write paths unprotected. References GDPR Article 32 (Security of processing) and SOC 2 Trust Services Criteria CC6.1 (Logical and Physical Access Controls) for the least-privilege alignment. Triggers on Supabase table policy authoring (CREATE POLICY, ALTER TABLE ... ENABLE ROW LEVEL SECURITY references). OWASP ASI03 (Identity & Privilege Abuse) defense.
fastapi-dependency-injection
by MILTONADINA
FastAPI dependency-injection patterns and anti-patterns -- request-scoped auth tokens leaking into singleton DI graphs, missing Depends() declarations on auth-required routes, dependency_overrides mis-use in tests bleeding into production code paths. Triggers on FastAPI route authoring (@app.get, @app.post, Depends() references). OWASP ASI03 (Identity & Privilege Abuse) defense for auth-token handling via DI.
rust-cargo-audit
by MILTONADINA
Cargo dependency auditing discipline -- cargo audit invocation in CI, cargo-deny advisory enforcement, RUSTSEC ID monitoring, and pinned-version vs floating-range tradeoffs. Triggers on Rust dependency authoring (Cargo.toml edits, Cargo.lock updates). OWASP AST08 (Update tampering) defense -- pinned versions plus CI advisory enforcement close the supply-chain side of the dependency-bump attack surface.
stripe-pci-scope-minimization
by MILTONADINA
Minimize PCI DSS scope when integrating Stripe -- token-only flows (Elements, Checkout, PaymentSheet), avoid raw PAN handling, audit-log discipline for cardholder-data adjacencies. References PCI DSS v4 Requirement 3 (Protect Stored Account Data) and Requirement 10 (Log and Monitor Access). Triggers on Stripe payment-intent / charge / customer code authoring. OWASP ASI03 (Identity & Privilege Abuse) defense for PCI-scoped credential handling.
nextjs-check-route-types
by MILTONADINA
Type-system hygiene for Next.js App Router routes -- any-typed params, missing route handler signatures, generated-types staleness from next dev and next build typecheck gaps. Triggers on Next.js route handler authoring (app//route.ts, app//page.tsx typed-segments). Not agent-specific security; complementary to runtime hardening.
nextjs-server-action-safety
by MILTONADINA
Defend against client/server boundary confusion in Next.js Server Actions. Patterns covered include authn drift in token-bound mutations, input validation at the action boundary, and mutation idempotency under retry. Triggers on Server Action authoring (functions with use server directive) and Next.js form-action references. OWASP ASI02 (Tool Misuse) plus ASI03 (Identity & Privilege Abuse -- authn drift) defense.
stripe-webhook-idempotency
by MILTONADINA
Stripe-specific webhook idempotency -- signature verification with the constructEvent API, retry-id reuse via event.id, double-fulfilment guards on PaymentIntent / Charge / Invoice events. Extends the universal security/webhook-idempotency skill with Stripe's event-id discipline and signing-secret rotation patterns. Triggers on Stripe webhook handler authoring (stripe.webhooks.constructEvent references, Stripe-Signature header handling). OWASP ASI02 (Tool Misuse) defense.
webhook-idempotency
by MILTONADINA
Defend against double-processing of webhook events across every provider (Stripe, GitHub, Slack, Twilio, SendGrid). Idempotency-key pattern + replay-window enforcement + dedupe-storage strategy. Triggers on any webhook handler authoring (route matching /webhook(s)?/, Stripe-Signature / X-Hub-Signature-256 header references). Defense against OWASP ASI02 (Tool Misuse) -- a replayed webhook that re-triggers a destructive tool is the canonical control-flow-level replay attack. Universal across providers; stack-specific signature verification lives in stack-specific skills.
cost-attribution
by MILTONADINA
Track per-client and per-feature LLM and cloud spend with cryptographic audit trail. Use whenever billable operations occur - LLM API calls, infrastructure provisioning, batch jobs. Required for any project with multiple clients or cost-sensitive operations. Integrates with Stratum's append-only billing_records table.
semgrep-scan
by MILTONADINA
Run Semgrep static analysis to detect insecure code patterns - SQL injection, XSS, path traversal, weak crypto, broken auth, OWASP Top 10 issues. Use on every file change touching security-sensitive code and as part of the tiered security gate. Complements gitleaks (which catches secrets) by catching insecure logic.
spec-extraction
by MILTONADINA
Convert a user's prose requirements into EARS-formatted specifications under /specs/. Use whenever the user describes new functionality, requirements, or a feature without an existing spec. Required first step for any non-trivial implementation - no spec, no work. Triggers on phrases like "add feature X", "build a Y", "we need to do Z", "the system should...".
proof-of-work
by MILTONADINA
Produce structured proof artifacts for every completed claim. Use this skill whenever the agent claims to have implemented something, tested something, deployed something, or completed a task. Every claim must ship with proof — git SHA, files changed, test command, exit code, output tail, and a reproducibility hash. Claims without proof are rejected by the claim-validator. Triggers on any output that asserts completion.
ears-spec-writing
by MILTONADINA
Write requirement specifications in EARS (Easy Approach to Requirements Syntax) notation. Use whenever authoring or editing files in /specs/, converting prose requirements into structured form, or reviewing existing specs for ambiguity. EARS uses five sentence patterns to produce unambiguous, testable, machine-readable requirements. Used by NASA, Airbus, Rolls-Royce, and Amazon Kiro IDE.
gitleaks-scan
by MILTONADINA
Run gitleaks against the working tree to detect committed secrets, API keys, tokens, and credentials. Use whenever a file has been written, before any git commit, before any push, and as part of the tiered security gate. Catches the most common cause of credential leaks - accidental commits.
multi-tool-failover
by MILTONADINA
Manage session continuity when switching between Claude Code, Codex CLI, Cursor, Antigravity, Kiro, or local LLMs (OpenCode, Aider). Use when approaching session limits, switching to a different tool's strengths, or when one tool fails and another must continue. Relies on the baton protocol and universal SKILL.md format.
session-summary
by MILTONADINA
Generate the end-of-session human-reviewable report. Use at session end before writing the baton. Aggregates all verified claims, lists open blockers, links proofs, and surfaces low-confidence items for human review. Required for any session that produced verifiable work.