TILLDEV DOCUMENTATION
Pick a product. Or read across.
Each product has its own quickstart, SDK references, and operational guides. The cross-product section covers things that look the same everywhere — auth, security, API conventions.
TillPulse
Observability for any project — SDKs, dashboard, alerts, deployment.
TillAuth
Auth + sessions + SSO. Passkeys, MFA, OAuth — first-class on every plan.
TillCache
Managed KV, durable queues, pub/sub — one HTTPS endpoint, one token.
- QuickstartNamespace, token, SDK — your first KV command, job, and pub/sub event.
- KVioredis-compatible get/set/incr/hset, TTLs, pipelines, raw command().
- QueuesDurable at-least-once work: push/pull/ack, visibility timeouts, a DLQ.
- Backends & vendorsUpstash, Durable Objects, or bring-your-own — chosen per namespace.
TillShield
Reactive security. Detection, rules, incidents, threat intel — part of TillDev.
- How detection worksThe security types TillShield reads, scored by severity and confidence.
- Writing rulesTriggers and the four actions — raise, quarantine, revoke, alert.
- Writing TillTell rulesDetection of your own: the format, tests, limits, and a backtest on recent events.
- TillHuntSearch your security events in the rule grammar, group and pivot, and turn a hunt into a rule.
- EndpointsThe TillTell agent for Linux, macOS and Windows: signed installs, approval by fingerprint, what each OS reports.
- IncidentsFindings grouped into cases, one timeline, merges, and the rule backlog.
- TillDrillTest what you own from the outside: prove ownership, run OWASP-tagged passive checks, measure latency by region, run capped load drills, and scan your app builds against MASVS.
- PostureConnectors that read your workspace and cloud accounts on a schedule, with no stored cloud keys, the inventory they keep, the checks that judge it, and drift from a baseline.
- VulnerabilitiesEverything the scanners find in one list, with a due date per level, owners, accepted risk, time to fix, and pull requests that upgrade a vulnerable dependency.
- Vulnerability disclosureA public page where researchers report to you, a sealed inbox with reply and decision clocks, rewards, and security.txt.
- EngagementsPurple-team exercises, red-team runs and pentests: rules of engagement, an owner’s authorisation, plays judged against what TillTell saw, tester links, retests and the report.
- Threat intelTor exits and addresses other workspaces flagged on every security event, and known-exploited CVEs and EPSS scores in the inventory.
- Insider riskTillTell rules for your own members and service credentials, with a score per person that is a reason to look, never a verdict.
- Compliance evidenceThirty controls judged from what your products already record, mapped to SOC 2, ISO/IEC 27001 and NIST CSF, with attestations and hash-chained snapshots an auditor can verify.
- Threat intel & privacyCross-customer sha256 indicators, anonymized and opt-out.
TillSecrets
Encrypted config store + dynamic secrets. Versioned, envelope-encrypted, pulled at boot.
- QuickstartProject, environment, token — set a secret and pull it into your app.
- Config storeEnv-var keys, immutable versions, rollback, service tokens, the SDKs, sync targets.
- Dynamic secretsLease short-lived Postgres credentials that auto-revoke on expiry.
- SecurityPer-org envelope encryption, an append-only audit log, SSRF-guarded backends.
TillArk
Backup, disaster recovery & failover — with restore-verification you can prove.
- QuickstartRegister a source and target, set a policy, take your first backup.
- Backups & retentionSchedules, GFS retention, deduplication, and cross-vendor anti-affinity.
- FailoverWarm-standby replicas, fenced promotion, and fail-back with observed RPO/RTO.
- Security & verificationSandboxed restore-verification, four-eyes approval, envelope-encrypted credentials.
TillForge
Sovereign, security-first code hosting — secrets can’t leak past the push boundary, and no mistake is fatal.
- QuickstartCreate a repo, mint a tfp_ token, clone and push over HTTPS from any editor.
- RepositoriesCreate, clone, and push over HTTPS + SSH; browse code, history, and blame.
- Branch protectionProtected default branch, secret-reject-at-push, one-click reflog recovery.
- SecuritySovereignty, the push boundary, AI egress governance, and recovery.
- Code scanningInjection, broken TLS, weak crypto and mobile storage flaws in every push; warn or block.
- IaC scanningTerraform, CloudFormation and Kubernetes judged by the posture checks before anything deploys.
- Build provenanceSigned SLSA v1 provenance for every CI build, with P-256 and ML-DSA-65; check any file offline.
- Source trackEvery protected-branch update graded against the SLSA Source track, signed as provenance and a verification summary.
- Supply chainWhich checks cover every repository, and deploys held for approvals and failing scans.
TillNotary
A tamper-evident transparency log and timestamping notary — signed inclusion proofs and witness-cosigned checkpoints you verify yourself, without trusting the operator.
- QuickstartCreate a log, pin its keys, submit a hash, and verify the receipt end to end.
- Inclusion & consistencyThe two proofs that make “append-only” a claim you can check, not a promise.
- Witnesses & refusalsIndependent co-signers, the k-of-n threshold, and what a refusal means.
- Evidence bundlesA portable proof file that verifies with zero network — for auditors and courts.
Cross-product
Reference material that spans every product in the family.
FOR LLMSOnboarding an AI agent? Point it at
tilldev.dev/llms.txt — a flat link map of every doc page with one-line summaries.Open llms.txt ↗