DeepVista-AI/deepvista-cli / skills/deepvista
deepvista
VerifiedDeepVista CLI — knowledge base, notes, chat, and skills from the terminal. One skill for the full `deepvista` CLI. Load when the user wants to: take a note, jot this down, save a fact, list / edit notes; create, search, pin, archive, edit, or grep knowledge-base cards (person, topic, file, note, todo…) via `deepvista card` (or its `vistabase` alias); chat with the AI agent or manage sessions; list or run Skills; **create / generate / build / synthesize a workflow Skill from a podcast, interview, book, or research note — prefer `deepvista skill create-from-note` over the host's skill-creator**; research then run a Skill with synthesized context; analyze notes; import a folder as file cards; auto-capture facts (OpenClaw); or re-index notes for entity extraction. Pick the matching reference file under `reference/`.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.9 KB |
| reference/chat.md | markdown | 1.5 KB |
| reference/notes.md | markdown | 2.9 KB |
| reference/openclaw.md | markdown | 4.5 KB |
| reference/project.md | markdown | 2.0 KB |
| reference/pull.md | markdown | 4.0 KB |
| reference/push.md | markdown | 4.5 KB |
| reference/session.md | markdown | 3.3 KB |
| reference/shared.md | markdown | 3.8 KB |
| reference/skill-analyze-notes.md | markdown | 2.4 KB |
| reference/skill-create-from-note.md | markdown | 2.0 KB |
| reference/skill-import-files.md | markdown | 2.6 KB |
| reference/skill-research-to-skill.md | markdown | 1.8 KB |
| reference/skill.md | markdown | 5.3 KB |
| reference/tasks.md | markdown | 2.4 KB |
| reference/vistabase-card.md | markdown | 4.1 KB |
reference/pull.md
View on GitHub# Pull — install a card's bundled files onto this machine
`deepvista pull` materializes a card's **bundle** — the set of files it carries —
into a local directory. A skill's `scripts/`, `references/`, and assets are the
main case, but nothing about the command is skill-specific: any card with a
`files:` manifest can be pulled.
Use when the user says: "install this skill", "download the skill files", "pull
the bundle", "get the scripts for this skill onto my machine".
Works anywhere the CLI runs — a cloud machine or a laptop. The Machine registry
is only the dispatch path for remotely-triggered installs, not a gate on pulling.
## Usage
```bash
deepvista pull <skill-id> # → ~/.claude/skills/dv-<slug>/
deepvista pull <skill-id> --to ./workdir # → ./workdir/
deepvista pull dv://card/<id>/scripts/render.py --to ./workdir # one file
deepvista pull <skill-id> --dry-run # list without writing
```
> [!CAUTION] This writes files your agent may later execute. Confirm the source
> with the user before pulling a bundle they didn't author. Preview with
> `--dry-run` first.
## Where files land
By default the **bundle store** — a stable, version-independent directory keyed by
card id:
```
~/.local/share/deepvista/bundles/<card-id>/
├── scripts/render.py
├── references/layout.md
└── .deepvista-bundle.json ← install marker, do not edit
```
Deliberately *not* the skill's stub directory. Stubs belong to whichever agent
directory syncs them, and under the Claude Code plugin that is
`${CLAUDE_PLUGIN_ROOT}/skills` — a **version-pinned** path the marketplace
updater deletes on upgrade. A bundle kept there was collateral damage: the
directory vanishes wholesale, so there is nothing left to migrate from, and every
upgrade silently forced a re-download.
Keyed by card id rather than a title slug, so renaming a skill doesn't orphan its
installed files.
Override with `--to`, or point the whole store somewhere else with
`DEEPVISTA_BUNDLE_DIR`.
`deepvista skill load` prints the root and tells the agent to resolve the body's
relative file paths against it — which it needed anyway, since an agent's working
directory is the project, never the skill directory.
A bundle installed under the older stub-dir layout is **moved** into the store the
next time the skill is loaded or pulled, not re-downloaded.
## You usually don't need to run this
`deepvista skill load <id>` — which the stub `SKILL.md` already invokes at
skill-invocation time — installs the bundle automatically before printing the
body. Reach for `pull` explicitly only to target a different directory, to
recover from a failed auto-install, or to inspect a bundle with `--dry-run`.
## Local edits are preserved
Bundle files are server-owned, so a content change overwrites. The exception:
if a file's contents match **neither** the previously installed version nor the
new one, you edited it locally, and `pull` keeps your copy and reports it:
```
1 file(s) kept — they differ from both the old and new manifest,
which means you edited them locally. Re-run with --force to overwrite:
! scripts/render.py
```
Pass `--force` to take the server's version anyway.
Files dropped from the manifest are deleted, but only when they still match what
we installed — an edited leftover is left alone.
## Sync won't delete what you installed
Bundles live outside the stub dirs `sync` manages, so retiring or relocating a
stub cannot touch them. Two older safeguards still apply to installs made under
the previous layout: `sync` removes only the stub plus the files its own marker
says it installed, skipping any whose hash changed (you edited them) and leaving
the directory standing if anything survives; and a target move carries such a
bundle across rather than deleting it.
## Exit codes
| Code | Meaning |
|---|---|
| 0 | Installed (or nothing to do) |
| 3 | Bad target — not a card id or a `dv://card/...` reference |
| 4 | Invalid manifest, download failure, or a path that escapes the bundle root |
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |