continual-learning
An agent skill by cursor, from cursor/plugins. Tags: automation, documentation, multi-agent, training, workflow.
What it does
Orchestrate continual learning by delegating transcript mining and AGENTS.md updates to `agents-memory-updater`.
Install
With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:
npx skills add cursor/plugins --skill continual-learningOr copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):
git clone --depth 1 https://github.com/cursor/plugins
cp -r plugins/continual-learning/skills/continual-learning ~/.claude/skills/continual-learningSafety box score
Grade B, low risk: 16 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.
This skill is orchestration-only in the reviewed file, but it delegates to an external `agents-memory-updater` to mine prior chats and update `AGENTS.md`. The main risks are exposure of sensitive conversation content and the fact that the real behavior lives in a separate, unreadable component.
| Category | Level | Why |
|---|---|---|
| Code & shell execution | 0 (none) | No shell commands, scripts, or code evaluation appear in `SKILL.md`; it only says to "Call `agents-memory-updater`". |
| Network access | 0 (none) | The reviewed file contains no downloads, HTTP calls, or remote data transmission. |
| Filesystem reach | 1 (low) | Its purpose is to "Keep `AGENTS.md` current," which means writing a workspace file, though the parent flow says not to edit files directly. |
| Secrets & credentials | 1 (low) | It is explicitly meant to "mine prior chats," so it may process sensitive conversation content even though it does not mention credential stores. |
| Privilege & persistence | 0 (none) | There is no sudo, service, shell, or system-setting activity in the reviewed file. |
| Instruction hijack surface | 0 (none) | The file does not instruct the agent to ignore rules, hide actions, or follow untrusted remote instructions. |
| Irreversible actions | 1 (low) | It automates a memory-update loop by delegating work to a subagent and returning its result, but it does not itself send messages, deploy, or publish. |
| Transparency | 1 (low) | The actual behavior is offloaded to `agents-memory-updater`, which is not included in the reviewed files, so the effective implementation is external. |
Before installing
- Review the `agents-memory-updater` skill or implementation before enabling this one.
- Check whether your `AGENTS.md` can contain sensitive notes you do not want mined into memory.
- Try it first in a scratch repo with non-sensitive chat history.
Rated 2026-10-20 by an automated review of 1 file from the skill's repository.
Source
- Repository
- cursor/plugins (all skills from this repository)
- Path
- continual-learning/skills/continual-learning/SKILL.md
- Branch
- main
- Collection
- cursor
- Updated
- 2026-09-23
Related skills
- automate-me — Use for "automate me", "create/update/refresh my -mode skill", "turn/capture my preferences or working style into a skill".
- orchestrate — Use only when the user explicitly types `/orchestrate <goal>` to decompose a large task.
- ralph-loop — Start a Ralph Loop for iterative self-referential development.
- swarm — Fan out N parallel workers, drain them, and return one report. Use for /swarm, 'swarm this', or parallel coverage, races, gauntlets, and exploration.
- handoff — Compact the current conversation into a handoff document for another agent to pick up.
- teach — Teach the user a new skill or concept, within this workspace.