Tools Constellation
22 models, agents, media generators, clouds and design tools — 12 in production on engagements, the rest under evaluation or in sandboxed R&D. Each entry says what we use it for, not what the vendor says it does.
Index
Filter by category or maturity, or search by what a tool is used for. Every card has a stable anchor, so /tools#tool-claude links straight to it.
Category
Maturity
Showing 22 of 22 tools
Foundation Models · Coding · Agents · Data
Our primary model for agentic coding, spec drafting, long-context document review and structured extraction, called through private endpoints with every prompt hashed to the ledger.
Product page for Claude (opens in a new tab)Agents · Coding
We package our review commands, hooks and connectors as versioned plugins so every engineer's coding agent runs the same governed toolchain.
Agents · Coding
We encode house procedures — eval runs, landing-zone reviews, ledger export — as skills the agent loads only when the task calls for them, keeping context budgets tight.
Coding
The editor our engineers use for agentic implementation against a written spec, with every generated diff reviewed hunk by hunk before commit.
Product page for Cursor (opens in a new tab)Coding
Repository-scoped rule files, reviewed like code, that pin architecture boundaries, banned APIs and secret-handling patterns into every Cursor session.
Agents · Coding
Agent tooling we are running through our coding-agent harness on internal repositories to decide whether it earns a place in the build pipeline.
Video/Image
The generation API we use to orchestrate video models — this site's hero backdrop was produced through it with Kling and Seedance.
Product page for Higgsfield AI (opens in a new tab)Cloud · Foundation Models · Data
Managed model access inside client AWS accounts, reached over VPC interface endpoints so prompts and grounding data stay inside the account boundary.
Product page for Amazon Bedrock (opens in a new tab)Cloud · Foundation Models · Data
Model deployments and evaluations inside client Azure tenants, fronted by private endpoints and Entra ID so access follows the tenant's existing identity controls.
Product page for Microsoft AI Foundry (opens in a new tab)Foundation Models
An open-weight model line we benchmark in our eval harness for scenarios where inference has to run self-hosted, inside a boundary no external API can reach.
Agents
An open-source agent runtime we study in an isolated sandbox to understand tool-permission and prompt-injection behaviour before anything similar touches client systems.
Foundation Models
A second model for drafting and cross-model review: we run it against the same golden sets as Claude and route work to whichever scores higher for the task.
Product page for ChatGPT (opens in a new tab)Video/Image
Video model we used, via Higgsfield, to generate this site's hero backdrop; we use it for motion studies and explainer footage.
Video/Image
Video model we used alongside Seedance, via Higgsfield, to generate this site's hero backdrop and ambient motion loops.
Video/Image
Image model wired into our media pipeline as the keyframe generator; it is not yet enabled on our API account, so keyframes route through other models today.
Foundation Models
A model we score on our eval harness for reasoning and retrieval-heavy tasks before we would route any client workload to it.
Product page for Grok (opens in a new tab)Agents
Conversational agent tooling we explore in a sandbox to compare agent behaviour and guardrail responses across model providers.
Coding · Design
A browser-based app generator we use for disposable prototypes, so stakeholders can react to a working flow before we write the spec.
Product page for Bolt (opens in a new tab)Coding · Cloud
A hosted workspace with an AI agent that we use for throwaway proofs of concept — never for client production code or client data.
Product page for Replit (opens in a new tab)Coding · Data
An app builder we are evaluating for internal tools that a client's business team could own and maintain after hand-off.
Coding
An AI build tool we are assessing against our spec-driven workflow; it has no production role yet.
Design
Figma's built-in AI features, which we use to explore layout variants and draft prototype flows before a design review — the reviewed file, not the suggestion, is the source of truth.
Product page for AI Figma (opens in a new tab)Evaluation rubric
A tool moves from R&D to Evaluating to Production only by clearing the same five gates. Nothing is grandfathered: model and vendor updates re-run the rubric.
01 · Security review
We map where prompts, files and credentials go, check SSO and SCIM support, audit-log export and token scoping, and red-team the tool with our prompt-injection suite before it sees anything but synthetic data.
02 · Data residency
We confirm processing and storage regions, retention and training-use terms in writing, and prefer deployments inside the client's own tenancy reached over private endpoints.
03 · Eval-harness scores
Each candidate runs the same versioned task suites — coding, extraction, reasoning, refusal behaviour — and must beat or match the incumbent within a declared tolerance band on every suite.
04 · Cost
We measure tokens, latency and retries on the eval suites and project cost per completed task at the client's volume, including the human review time the tool's error rate implies.
05 · Exit path
Every tool sits behind our orchestration layer's provider interface, and prompts, rules and eval sets live in the client's repository — so swapping a vendor is a config change plus an eval run.
We run your candidate tools through the same rubric — security review, residency, golden-set scores, unit cost and exit path — and hand you the evidence, not a vendor ranking.