← 返回 Skills 市场
ironystock

ClawCast

作者 ironystock · GitHub ↗ · v0.2.6 · MIT-0
cross-platform ⚠ suspicious
294
总下载
0
收藏
0
当前安装
4
版本数
在 OpenClaw 中安装
/install claw-cast
功能描述
Bootstrap and automate OBS scenes for local or remote instances via agentic-obs + mcporter. Includes optional explicit target-switch step that writes agentic...
使用说明 (SKILL.md)

ClawCast

Use this skill to create a generic, reusable OBS automation baseline.

Prerequisites

  • mcporter installed/configured with obs MCP server
  • OBS WebSocket enabled on target OBS host (default 4455)
  • python3, sqlite3, ss (iproute2), and standard shell tools

Prerequisite 0 (required): verify obs MCP path

mcporter list
mcporter call 'obs.get_obs_status()'

If this fails, stop and finish mcporter + OBS MCP configuration first.

Workflow

  1. Switch target OBS host
  2. Start overlay HTTP server
  3. Rebuild baseline scene pack from skill assets
  4. Optionally apply transition + audio defaults
  5. Run recording smoke walkthrough
  6. Optionally run stream dry-run

Commands

# 1) Target OBS host (explicit write acknowledgement + DB path required)
./skills/clawcast/scripts/obs_target_switch.sh \x3Cobs-host-ip> 4455 \
  "$HOME/.agentic-obs/db.sqlite" --allow-cross-component-write

# 2) Start/verify overlay host server (serves skill directory only)
./skills/clawcast/scripts/start_overlay_server.sh

# 3) Rebuild baseline scenes + overlays
./skills/clawcast/scripts/rebuild_scenes.sh

# 4) Apply transition preset
./skills/clawcast/scripts/apply_transition_preset.sh Fade 300

# 5) Optional audio baseline
# export OBS_AUDIO_INPUTS="Mic/Aux,Desktop Audio"
./skills/clawcast/scripts/apply_audio_baseline.sh

# 6) Run walkthrough recording
./skills/clawcast/scripts/smoke_test_walkthrough.sh

# 7) Optional stream dry-run
./skills/clawcast/scripts/stream_dry_run.sh 15 "Intro" "Main Live"

Notes

  • This skill is boilerplate-first; customize scenes and assets after bootstrap.
  • Baseline-required overlays live in assets/overlays/.
  • Optional project-specific examples live in examples/project-specific/.
  • Avoid /tmp for persistent assets.
  • For remote OBS, prefer HTTP browser source URLs over file://.

Security & transmission

  • No API keys/tokens are required by default for this skill.
  • OBS control uses the configured WebSocket endpoint (default port 4455).
  • Overlay pages are served over local HTTP (:8787) and should remain on trusted LAN/VPN.
  • start_overlay_server.sh serves only the skill directory (not workspace root).
  • obs_target_switch.sh performs a cross-component config write only when given an explicit DB path argument plus --allow-cross-component-write.
  • Do not expose OBS WebSocket or overlay HTTP ports publicly.

References

  • references/scene-map.md
  • references/networking.md
  • references/troubleshooting.md
  • references/v0.2-features.md
安全使用建议
This skill appears to do what it says: bootstrap OBS scenes and serve local overlays. Before installing/running: - Review and back up any agentic-obs DB you might pass to obs_target_switch.sh. That script will update obs_host/obs_port in the specified sqlite DB if you provide the DB path and the --allow-cross-component-write flag. - Only run obs_target_switch.sh when you intentionally want to change the agentic-obs target; if unsure, skip that step. Verify the DB path points to the intended database (do not pass arbitrary files). - start_overlay_server.sh uses python3 -m http.server to serve the skill folder on the chosen port (default 8787). Confirm the machine/network are trusted and do not expose that port to the public internet. - Ensure mcporter is properly configured and that you trust the configured 'obs' MCP server before running any scripts that call it. - If you have low tolerance for configuration changes, run the scripts in a non-production/sandbox environment first and inspect scripts (they are plain shell) to confirm behavior.
功能分析
Type: OpenClaw Skill Name: claw-cast Version: 0.2.6 The skill bundle provides OBS automation but contains high-risk behaviors and vulnerabilities. Specifically, 'scripts/obs_target_switch.sh' performs a cross-component write to an external SQLite database ($HOME/.agentic-obs/db.sqlite) to configure connection parameters; while documented and guarded by a flag, the script lacks input sanitization, creating a potential SQL injection vulnerability. Additionally, 'scripts/start_overlay_server.sh' executes a background Python HTTP server to serve local assets. While these actions align with the stated purpose of bootstrapping OBS scenes, the combination of external file modification and insecure command construction warrants a suspicious classification.
能力评估
✓ Purpose & Capability
Name/description (OBS automation via mcporter/agentic-obs) align with required binaries (mcporter, python3, sqlite3, ss) and the included scripts. The scripts perform expected tasks: start a local overlay HTTP server, create browser sources and scenes via mcporter, run smoke tests, and optionally write an agentic-obs DB entry.
ℹ Instruction Scope
Most runtime instructions are narrowly scoped to OBS automation (mcporter calls, local HTTP server serving the skill folder, scene creation). One instruction/script (scripts/obs_target_switch.sh) will update another component's sqlite DB (agentic-obs config) when given a DB path and the explicit --allow-cross-component-write flag; this is intentional and documented, but it modifies other component configuration and therefore warrants explicit operator attention before use.
✓ Install Mechanism
This is an instruction-only skill (no remote download/install). All code is included in the repo as scripts and static overlays; there is no network install step or external archive extraction. The only runtime network behavior is starting a local python http.server to serve the skill directory.
ℹ Credentials
The skill requests no credentials or env vars by default. Optional envs (OVERLAY_PORT, OVERLAY_BASE_PATH, OBS_AUDIO_INPUTS, MIC_MUL, DESKTOP_MUL) are reasonable for the described tasks. However, the skill can be used to update an agentic-obs sqlite DB when the operator supplies a DB path; the registry metadata lists no required config paths even though the cross-component write is a supported action — this should be considered before granting filesystem access or passing an agentic-obs DB path.
✓ Persistence & Privilege
The skill does not request 'always: true' and is user-invocable. It writes only to its own skill directory (served by the HTTP server) and only writes to another component's DB when explicitly invoked with a DB path and acknowledgement flag. It does not modify other skills' configuration by default.
如何使用
  1. 确保已安装 OpenClaw（本地或 Docker 部署）
  2. 在对话框中输入安装命令：/install claw-cast
  3. 安装完成后，直接呼叫该 Skill 的名称或使用 /claw-cast 触发
  4. 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v0.2.6
- Added metadata section to SKILL.md with OpenClaw emoji, homepage link, and required binaries. - No changes to skill commands, workflow, or usage instructions. - Documentation now includes structured metadata for improved integration and discoverability.
v0.2.5
claw-cast 0.2.5 - Updated obs_target_switch.sh: now requires database path as an explicit argument and uses a new --allow-cross-component-write option. - Improved security by requiring explicit acknowledgment for cross-component config writes. - Documentation in SKILL.md and README.md updated to reflect new usage and safety steps. - No changes to core automation or workflow; base feature set and commands remain stable.
v0.2.4
- Added explicit prerequisite to verify mcporter OBS MCP connectivity before use. - Strengthened command examples: obs_target_switch.sh now requires ALLOW_CROSS_COMPONENT_WRITE=1 and AGENTIC_OBS_DB, providing safer cross-component config control. - start_overlay_server.sh limits server scope to the skill directory for improved security. - Expanded security and transmission documentation, highlighting changes to overlay/server file serving and configuration environment variables. - New script require_obs_mcp.sh checks OBS MCP availability before usage. - Documentation updates: improved step breakdowns, warnings, and prerequisites in README.md and SKILL.md.
v0.2.3
claw-cast v0.2.3 - Initial public release of the ClawCast skill for OBS automation - Provides scripts to bootstrap and automate baseline OBS scenes - Supports smoke test workflows: recording and stream dry-runs - Asset organization and overlay hosting guidance included - Emphasizes LAN/VPN safety and boilerplate-first customization - Comprehensive workflow and command documentation in SKILL.md
元数据
Slug claw-cast
版本 0.2.6
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 4
常见问题

ClawCast 是什么？

Bootstrap and automate OBS scenes for local or remote instances via agentic-obs + mcporter. Includes optional explicit target-switch step that writes agentic... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 294 次。

如何安装 ClawCast？

在 OpenClaw 或 Claude Code 对话框中运行命令「/install claw-cast」即可一键安装，无需额外配置。

ClawCast 是免费的吗？

是的，ClawCast 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。

ClawCast 支持哪些平台？

ClawCast 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。

谁开发了 ClawCast？

由 ironystock（@ironystock）开发并维护，当前版本 v0.2.6。

推荐 Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 留言讨论