Moonsail

Orient First#

Moonsail is an agent-run media workspace and production control plane. Spaces hold a brief, immutable media Objects, and Jobs. The MCP endpoint lets agents select a curated model, generate media, keep provider outputs under a content hash, and review the stored bytes.

The skills exist to teach production judgment: how to turn source truth into a specific treatment, choose an execution path, inspect the actual result, and repair weak work. Moonsail supplies the live tools and accountable runtime; skills supply the standards for using them well. The goal is high-quality, source-aware media rather than generic generation or opaque automation.

Generations spend provider credits. Before a paid Job, inspect live pricing, estimate variants and repair reserve, and get explicit approval. Pass that ceiling as maxCostMicrousd. The prototype has no credit purchase tool. Never treat permission to make media as permission to increase the approved spend.

Architecture#

Every production uses three layers: the shared execution-and-evidence kernel, one or more focused content or operational skills, and a task-local treatment. Read Production Kernel for capability, policy, budget, provenance, evidence, admission, repair, and delivery rules. Keep scripts, storyboards, prompts, references, claims, timing, and budget decisions with the current job. Focused skills supply only the judgment needed to shape and evaluate them.

Establish Context#

• Read the user's request and the relevant Space brief, sources, references, and existing assets before making factual or creative assumptions.

• scripts/context.mjs --json can discover local configuration, context files, and installed skill versions. It is a discovery aid, not a production gate.

• Use get_object for stored source material and list_jobs to recover recent work. Read references/tool-belt.md when tool semantics are unclear.

• Ask only for missing information that materially affects safety, factual accuracy, or the requested result. Do not turn every job into a complete intake questionnaire.

Never invent product facts, brand rules, customer evidence, compliance status, or allowed claims to fill a gap.

Choose Relevant Guidance#

The table is discovery help, not a closed taxonomy. Combine skills when the artifact genuinely needs more than one kind of judgment, and continue with this root alone when no focused skill fits.

Choose an Execution Path#

• Inspect the operations and enabled models available now. Choose by required inputs, output modality, references, duration, aspect ratio, language, cost, speed, and fit for the specific artifact.

• Create one Job for one model operation. Keep multi-step production logic in the agent and store each durable result as an Object.

• Runtime validation owns model availability, compatibility, cost ceiling, and Job state. Follow its decision; do not duplicate those rules in prompts.

• Treat model and provider identities as live operational facts, not permanent rankings or user-facing creative language.

Shared Boundaries#

• Sources and claims: Trace factual product, brand, customer, legal, and compliance claims to supplied or Space sources. Distinguish a visible demonstration from an outcome the artifact does not prove.

• Identity and consent: Do not create or reuse a real person's likeness, voice, or persona without authorization appropriate to the requested use. Stop when consent or allowed use is unclear.

• Publishing and mutation: Do not publish, approve on a user's behalf, destructively change Space state, or silently expand scope or spend.

• Provenance: Preserve the model, prompt, references, Job ID, Object IDs, estimated cost, status, and output location recorded by the runtime.

• Honest state: Failed, incomplete, unavailable, or uninspected work cannot be reported as complete or creatively approved.

Review the Actual Artifact#

A prompt, brief, storyboard, Job metadata, or URL string is not a creative review. Give a modality-capable reviewer the actual final image, video, or audio artifact. For web content, inspect the exact source, DOM, metadata, links, and rendered page states available to the review.

• Create a review Job for the exact final Object ID, then poll it with get_job. For an external artifact, first copy it with put_object.

• Confirm the result's review receipt identifies the artifact by Object ID and SHA-256, reports the actual coverage, and is delivery-eligible. A blocked, degraded, mismatched, or unresolved review cannot be presented as a creative pass.

• Treat receipt validity and deterministic media probes as supporting evidence, not creative approval. Obvious visual failures such as extra limbs, impossible contacts, identity drift, contradictory action state, or broken continuity remain blockers even when the file decodes and hashes match.

• Inspect the exact final revision and enough of its spatial or temporal content to support the verdict.

• Name strengths worth preserving, then diagnose the most material observable cause of failure. Point to regions, frames, moments, or time ranges when the medium permits it.

• Repair the diagnosed cause without casually degrading work that already passes.

• After any repair, store and review the new Object. A pass on an earlier Object or SHA-256 does not transfer to the repair.

• The prototype does not support finding waivers. Report an accepted gap as unverified; do not change the receipt.

• At delivery, bind the output to the final review receipt. Its reviewed Object ID and SHA-256 must match the delivered Object exactly.

• If the reviewer cannot consume the modality or sufficient evidence, report review as blocked or degraded rather than passing it.

• Do not claim that the media analyzer reviewed text, DOM, links, or rendered web behavior. Use deterministic parsers and browser evidence for those artifacts, and keep unavailable states explicitly unassessed.

Deliver in the shape most useful to the request. Include final outputs, runtime-backed provenance, what was actually inspected, and any material gap; no universal prose template is required.

References and Helpers#

• references/production-kernel.md defines the universal execution, policy, evidence, admission, repair, and delivery contract.

• references/tool-belt.md describes Moonsail tools.

• references/verification.md contains objective checks that may be useful for the requested output.

• references/narrative-integrity.md separates causal story reality and believable blocking from artistic presentation.

• references/delivery.md is an optional structured handoff template.

• scripts/verify.mjs <files...> surfaces deterministic document findings. It does not judge creative quality or replace actual-artifact review.

• scripts/hook.mjs powers optional post-edit reminders and never blocks a production turn.