A field manual that makes smaller models more reliable. Retrieval cards cover retrieve-before-answer, compound questions, source units, exceptions, modal verbs, shown arithmetic, and what to say when the pack is silent. Includes the labeled Lantern Library example.
Explore Knowledge Images
Search and filter portable packs by format, integrity, license, and capability.
Red Seal house knowledge for L'Obscura: nine He lines, 77 colourways, size chart, fit protocol, and policy. Public storefront facts only. No customer measurements or supplier wholesale.
First-party intake Knowledge Image for the hiveforensics.com desk: identity, booking policy, Florida coverage, and FAQ. Not a customer corpus.
164 React scaffolding tracks for starting web, mobile, desktop, and full-stack apps. Each track has keywords, when-to-use notes, official CLI commands, key packages, and composition recipes. Seeded as a Knolo V5 knowledge image for lexical retrieval by humans and coding agents.
Design and operate dependable webhooks with signature verification, idempotency, retries, ordering, and replay-safe consumers.
A pragmatic guide to test layers, fixtures, contracts, end-to-end coverage, flake control, and release confidence.
A queryable guide to evolving JSON and event schemas, compatibility, enum changes, deprecation, and contract testing.
A practical guide to rate limiting, quotas, fairness, Retry-After, distributed counters, and graceful overload handling.
A field guide to logs, metrics, traces, SLOs, cardinality, sampling, redaction, and incident debugging.
A queryable guide to promises, the event loop, cancellation, concurrency, timeouts, streams, and async failure modes.
A practical playbook for feature flags, staged releases, targeting, kill switches, measurement, and cleanup.
A practical playbook for safe schema changes, expand-contract migrations, backfills, indexes, and rollback decisions.
A vendor-neutral playbook for debugging container images, startup failures, networking, permissions, signals, and resource limits.
A practical accessibility review guide for semantics, keyboard use, focus, forms, contrast, motion, and repeatable testing.
A queryable handbook for golden sets, LLM judges, pass@k, leakage, and prompt regression — written for people who ship prompts.
A queryable handbook for paging vs tickets, severity, flapping, runbooks, and handovers — written for people who get woken up.
A queryable handbook for major/minor/patch, pre-releases, caret vs tilde, and changelogs — written for people who ship libraries.
A queryable handbook for authorization-code + PKCE, redirect URIs, refresh rotation, and SPA token storage — written for people who ship login.
A queryable handbook for flexbox, grid, stacking, sticky, and container queries — written for people who ship UI, not spec excerpts.
A queryable handbook for B-tree, GIN, BRIN, partial, and covering indexes — written for people who wait on slow queries.
How to ground LLM answers in retrieved evidence, cite it, and abstain when the pack does not say — including prompt-injection boundaries.
Provider-neutral Terraform state, locking, backends, and workspaces — for teams that cannot afford a surprise terraform apply.
A deterministic customer-support policy for refunds, returns, digital goods, and damaged items — written to be queried by a support agent.
Practical application security for sessions, injection, secrets, and authz — written to be queried while you design or review.
A portable incident-response playbook: severity, commander, comms, mitigation first, and after-action — for people who are on call.
A queryable handbook for HTTP methods, status codes, errors, caching, and idempotency — written for people who ship JSON APIs.
How to mount a Knowledge Image, query it, and read a retrieval receipt — the real runtime contract, not a seed listing.
A portable handbook for undoing, recovering, and rewriting Git safely — with receipts, not folklore.