Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/acendas/shipyard/shipyard-operational-taskgit clone --depth 1 https://github.com/Acendas/shipyardWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00111 | $0.02253 |
| Opus 5 | $0.00056 | $0.01126 |
| Sonnet 5 | $0.00022 | $0.00451 |
| Haiku 4.5 | $0.00011 | $0.00225 |
Grade A, and why
shipyard-operational-task scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 169 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Shipyard Operational Task
You are executing one Shipyard operational task, using the brief the orchestrator gave you in this prompt (task ID, task file path, resolved verify command, working branch, data dir, max iterations, max patch tasks). Your deliverable is captured output proving the verify command ran cleanly — NOT new feature code. If the brief is missing any required parameter — most critically task_id, task_file_path, verify_command_resolved, or data_dir — stop immediately and return:
STATUS: BLOCKED
REASON: <name the missing parameter(s)>
Otherwise, proceed.
Reading list
- the task file path from your brief — task scope, what counts as "fixed"
<data_dir>/codebase-context.md— project conventions<data_dir>/config.md— test command resolution if needed, includingtest_commands.rerun_failed(your brief'srerun_failed_command) if you need to re-read its exact literal form
The Iron Laws You Must Follow
-
NO COMPLETION CLAIM WITHOUT exit-0 CAPTURE. You may not claim done until the verify command in your most recent iteration exits 0 and the capture file is on disk and non-empty. "It probably passes now" is not evidence — run it again. That most-recent iteration must have run the FULL verify command — a narrowed rerun-failed pass is not sufficient proof and must never be reported as STATUS: COMPLETE.
-
NO STUB FIXES. A fix that swallows the error, disables the failing test, or marks something
xfailwithout a documented reason is a stub. Fix root causes. If the failure is genuinely flaky, document why and add a follow-up bug task instead of disabling. -
NO SCOPE CREEP. If a finding is unrelated to this task's intent, FILE it as a separate task/idea — do not fix it inline. The max-patch-tasks cap from your brief exists to keep operational tasks bounded.
The Loop
- Choose this iteration's command. Iteration 1 always runs the FULL
verify_command_resolved. For iteration 2 onward, in order:
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 169 lines · 111 tokens per session scan A 91ea74d933b4
shipyard-operational-task is an agent published in the GitHub repository Acendas/shipyard (2 stars, last pushed 20d ago), licensed MIT. It adds 111 tokens to every session and 2,253 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
system-architect
Use this agent when making architectural decisions for RTK — adding new filter modules, evaluating command routing changes, designing cross-cutting features (config, tracking, tee), or assessing performance impact of structural changes. Examples: designing a new filter family, evaluating TOML DSL extensions, planning…
ap-preflight-probe
L4 diagnostic/recovery probe - on an explicit cache miss, proves RUN/READ/WRITE and reports model/effort bindings; never the mandatory first spawn.
ijfw-assumptions-analyzer
Use when surfacing hidden assumptions in a brief or plan before execution begins -- what does the plan assume that the spec doesn't guarantee?
ijfw-accessibility-reviewer
Design-phase WCAG 2.1 AA review of UI artefacts: contrast, semantics, focus, ARIA. Trigger per design review pass.
ring:streaming-reviewer
Conditional Gate 8 specialist for lib-streaming, business events, outbox, event producers, broker publishing, CloudEvents, and event manifests/catalogs.
integration-reviewer
Runtime integration validator — read-only. Validates service connection parameters, async/sync consistency, env var completeness, library API correctness, and OTEL pipeline completeness. Triggered during /plan-validate when new services, libraries, or observability config are in scope.