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 skills add lgbarn/shipyard --skill lessons-learnedgit clone --depth 1 https://github.com/lgbarn/shipyardWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/lgbarn/shipyard/lessons-learned)<a href="https://agentmods.dev/skills/lgbarn/shipyard/lessons-learned"><img src="https://agentmods.dev/badge/skills/lgbarn/shipyard/lessons-learned/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/lgbarn/shipyard/lessons-learned"><img src="https://agentmods.dev/badge/skills/lgbarn/shipyard/lessons-learned.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Prompt Injection · line 6 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
What 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 5.1 | $0.00079 | $0.01593 |
| Opus 5.5 | $0.00032 | $0.00637 |
| Sonnet 5.5 | $0.00016 | $0.00319 |
| Haiku 4.5 | $0.00008 | $0.00159 |
Grade A, and why
lessons-learned 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 6d 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 — 215 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lessons Learned
Iron Law
NO PHASE SHIPS WITHOUT LESSON CAPTURE
No exceptions:
- Not for "small phases" or "just a bugfix"
- Not for "we're in a hurry"
- Not for "nothing notable happened"
- Even smooth phases confirm what's working — capture it
When to Use
- After phase completion during
/shipyard:ship(Step 3a) - When reflecting on completed work to extract reusable knowledge
- When a build summary contains notable discoveries worth preserving
Natural Language Triggers
- "what did we learn", "capture lessons", "retrospective", "lessons learned"
When NOT to Use
- Mid-implementation (wait until the phase ships — partial lessons lack full context)
- Documenting code design decisions (put those in inline comments or architecture docs)
- Tracking active bugs (use issue tracker; lessons capture completed learning, not open problems)
Timing Rule
Run immediately after ship, before the next phase starts. Lessons captured 24+ hours later lose the "why" behind decisions. If you're about to start the next phase, capture lessons first.
Overview
The lessons-learned system captures discoveries, patterns, and pitfalls found during implementation and feeds them back into the project. Lessons are stored in .shipyard/LESSONS.md and optionally surfaced in CLAUDE.md so future agents benefit from past experience.
LESSONS.md Format
Store lessons in .shipyard/LESSONS.md using this exact structure:
# Shipyard Lessons Learned
## [YYYY-MM-DD] Phase N: {Phase Name}
### What Went Well
- {Bullet point}
### Surprises / Discoveries
- {Pattern discovered}
### Pitfalls to Avoid
- {Anti-pattern encountered}
### Process Improvements
- {Workflow enhancement}
---
New entries are prepended after the # Shipyard Lessons Learned heading so the most recent phase appears first. Each phase gets its own dated section with all four subsections.
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.
- 6d ago Changed · -6 tokens per session c3bbceda860b
- 1mo ago First seen · 215 lines · 85 tokens per session scan A dd5f2fabc880
lessons-learned is a skill published in the GitHub repository lgbarn/shipyard (65 stars, last pushed 6d ago), licensed MIT. It adds 79 tokens to every session and 1,593 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-09-03.
Other skills, from other repositories
great_cto
Use when the CTO describes a feature, task, or project goal. Orchestrates the full SDLC pipeline automatically based on project type.
opportunity-solution-tree
Build an Opportunity Solution Tree (OST) to structure product discovery — map a desired outcome to customer opportunities, possible solutions, and experiments. Based on Teresa Torres' Continuous Discovery Habits. Use when the team is unclear what to build next, when multiple opportunities compete, or before writing a…
done-blocked
Reusable reporting contract for any agent that hands work back to the pipeline. Forces ONE of two terminal statuses (DONE or BLOCKED) with a specific evidence shape. Stops vague "probably finished" and "kind of stuck" verdicts.
outcome-roadmap
Transform an output-focused roadmap (feature list) into an outcome-focused one. Rewrites initiatives as outcome statements reflecting user and business impact. Use when a roadmap lists features instead of results, when making a roadmap more strategic, or when communicating what success looks like vs what will be built.
pm-planning
Decomposition methodology for pm agent — turns an approved ARCH document into a Beads task list with explicit dependencies, time-boxes, and acceptance criteria. The pipeline can only orchestrate work it can see; this skill defines what "seeable work" looks like.
protocol-tracker
Internal skill for commands. The tracker verb-resolution protocol - how a lets-tracker block resolves through the active adapter, the neutral read shape, body crossing, degradation, preflight and trust. Loaded by a command or skill right before its first tracker verb. Do not trigger on user conversation.