Skill

fulcrum-write-memory-on-completion

Persist memory after any task involving decision, trade-off, or surprising finding. Applies after complete_agent_run when work produced durable knowledge.

Install

npx claudepluginhub moabualruz/fulcrum --plugin fulcrum

Tool Access

This skill uses the workspace's default tool permissions.

Preview

After completing a task — especially one with a decision, trade-off, or surprising finding — call `fulcrum action exec write_memory`. Makes Fulcrum L2 memory graph valuable over time. Every memory = gift to next agent (or future self).

Supporting Assets

agents/openai.yaml

SKILL.md

Similar Skills

cache-components

Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.

cache-components

139.1k

claude-opus-4-5-migration

2 files

Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.

claude-opus-4-5-migration

83.2k

claude-code-plugin-release

1 file

Automates semantic versioning and release workflow for Claude Code plugins: bumps versions in package.json, marketplace.json, plugin.json; verifies builds; creates git tags, GitHub releases, changelogs.

claude-mem

64.7k

Stats

Parent Repo Stars1

Parent Repo Forks0

Last CommitApr 20, 2026

Actions

View Source View Plugin View on GitHub View README

Write memory on completion

After completing a task — especially one with a decision, trade-off, or surprising finding — call fulcrum action exec write_memory. Makes Fulcrum L2 memory graph valuable over time. Every memory = gift to next agent (or future self).

When

Non-obvious architectural choice.
Discovered a gotcha (library quirk, flaky test, subtle bug).
Changed convention, introduced new pattern.
Task outcome worth remembering.
Just cited recalled memories — link output back.

Skip: trivial typo fixes, pure formatting, reverts of same-session work.

How

fulcrum action exec write_memory --json '{
  "workspace_id": "ws_123",
  "kind": "task_outcome",
  "content": "What happened and why it matters.",
  "tags": ["packages/core/src/memory", "FTS5", "fallback"]
}'

`kind`

task_outcome: what happened this task. "Did X to achieve Y; tests pass; PR #42 merged." Default for completion writes.
decision: architectural / convention choice. "Use text_pair tokenization for reranker — batching pairwise inputs avoids 40% throughput hit." Read-heavy over time.
lesson: something that generalizes. "SQLite FTS5 MATCH raises any SQLITE_ERROR, not specific subclass — catch broadly in fallback paths." Reusable across tasks.

`scope`

task: bound to task lifecycle. Use for task_outcome.
project: applies workspace-wide. Use for decision + lesson.

`tags`

Primary recall vector. Include:

File paths touched (packages/core/src/memory/recall.ts).
Component / concept names (FTS5, reranker, WIP limiter).
Error types / symptoms (SQLITE_ERROR, unterminated string).

Citing prior memories

Relied on recalled memories this session? Link in content: Supersedes M-0301. Builds on M-0423. Builds the graph CoS traverses when planning.

Red flags

Completed non-trivial run, no memory → knowledge gone when conversation ends.
Wrote memory with empty tags → never recalled; add paths + concepts.
lesson that is actually task_outcome → reclassify. Lesson must read useful to stranger with no context.