Install
$ agentstack add skill-yellowbee686-everything-analysis-agent-cbeta-research ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
CBETA Research
Overview
Use the mcp_servers/CbetaMCP submodule as the live tool server. This skill provides startup policy, tool selection, and citation workflows; it does not duplicate the MCP tool implementations.
Startup Policy
Do not start the MCP server when the skill is loaded. Start it only when the user task needs a live CBETA query.
- Check the endpoint first:
python3 skills/cbeta-research/scripts/check_server.py --url http://localhost:18765/mcp/
- If unavailable, start on demand:
skills/cbeta-research/scripts/start_server.sh
When running inside an agent shell that cleans up background processes, keep the server in a long-running foreground session:
CBETA_MCP_FOREGROUND=1 skills/cbeta-research/scripts/start_server.sh
- Re-run the check. If it still fails, report dependency, port, or server-log diagnostics.
The default MCP endpoint is http://localhost:18765/mcp/.
Windows Preference
On Windows, prefer running CBETA MCP commands inside WSL2. Fall back to PowerShell only when WSL2 is unavailable or the user explicitly asks for native Windows execution.
- If the agent is already running in WSL2, use the normal Linux commands above.
- If the agent is running in native Windows PowerShell or CMD and WSL2 is available, run Linux commands through
wsl.exe -- bash -lc. - If WSL2 is unavailable, use
scripts\start_server.ps1from PowerShell. - Convert Windows paths such as
C:\Users\name\repoto WSL paths such as/mnt/c/Users/name/repobefore calling Linux scripts.
Example from native Windows:
wsl.exe -- bash -lc 'cd /mnt/c/path/to/everything-analysis-agent && skills/cbeta-research/scripts/start_server.sh'
Fallback from native PowerShell:
.\skills\cbeta-research\scripts\start_server.ps1
Query Workflow
- Use
references/tool-map.mdto choose the smallest useful tool set. - For broad discovery, start with
cbeta_all_in_oneorcbeta_search_sc. - For exact Buddhist canon metadata, use catalog or work tools before content retrieval.
- For final answers, prefer source-backed snippets from
get_cbeta_lines,cbeta_kwic_search, orget_juan_html. - Include
work,title,juan, andlineheadwhen available. Preserve traditional Chinese source text unless the user asks for simplified Chinese.
References
- Read
references/tool-map.mdfor tool categories, common parameter shapes, and recommended flows. - Read
references/setup.mdfor submodule initialization, dependency installation, endpoint configuration, and troubleshooting.
Output Rules
- Distinguish search hits from verified source text.
- Show CBETA identifiers in citations:
work,title,juan,linehead, and URL fromcbeta_gotowhen available. - Report MCP startup failures with the check/start command and relevant log path.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: yellowbee686
- Source: yellowbee686/everything-analysis-agent
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.