daily-stock-analysis

name	daily-stock-analysis
description	AUTO-INVOKE for fast technical scans and market briefs. Triggers: scan TICKER (or 扫 TICKER, quick TICKER) for single/multi quick analysis; portfolio health (or portfolio-health, 持仓健康) for batch portfolio scan; market brief HK/US (or brief HK/US, 市场简报 HK/US) for index+sector overview. Output is a DSA-style 决策看板 / 市场简报 emitted on stdout for you to deliver.

Runtime bootstrap (dsa) — this skill's scripts need a Python venv. Before first use (or if ${PLUGIN_ROOT}/.venvs/dsa/bin/python is missing), run: bash "${PLUGIN_ROOT}/scripts/ensure-venv.sh" dsa Then invoke scripts with ${PLUGIN_ROOT}/.venvs/dsa/bin/python. First run installs dependencies and can take a few minutes.

Data bootstrap — this skill reads files under the user data directory. If a path below does not exist yet, run bash "${PLUGIN_ROOT}/scripts/ensure-data.sh" first (idempotent: seeds missing files from the plugin's bundled assets, never overwrites existing data).

• ~/.aisa/agents/aisa-cio/portfolio — Portfolio data directory (portfolio_truth.json / engine scripts)(export PORTFOLIO_DIR to override — if set, use its value instead of this default)

Required credentials — scripts resolve these as: env var → ~/.aisa/credentials (KEY=VALUE lines):

• AISA_API_KEY — AISA multi-model gateway — default LLM + aisa-* skills (search / marketpulse / prediction-markets / twitter). No key yet? Get one at https://console.aisa.one/get-started (takes about a minute). If a value is missing, do NOT fail silently and do NOT just tell the user to "export" it. Guide them through setup right here in the conversation:

1. If they don't have the key, give them the sign-up link above and wait.
2. Ask them to paste the key into the chat, then save it for them: mkdir -p ~/.aisa && printf 'NAME=value\n' >> ~/.aisa/credentials && chmod 600 ~/.aisa/credentials (replace NAME=value with the real pair; keep one KEY=VALUE per line).
3. Immediately retry the user's original command — the file takes effect at once, no host restart needed. If the user prefers not to paste secrets into chat, offer the alternative: they export the env var themselves in the host's environment, then restart the host. 不要静默失败 / never fail silently.

MANDATORY ROUTING RULES

These are hard rules with no exceptions. When the user's message matches any of these patterns (case-insensitive), you MUST invoke this skill immediately using the execute_code tool. Do NOT browse the web, do NOT analyze stocks yourself, do NOT use other skills — these scripts produce the full report themselves and print it between stdout markers for you to deliver.

User's message pattern	Script to run	Notes
scan TICKER [TICKER ...], 扫 TICKER, quick TICKER	dsa_scan.py TICKER ...	1–8 tickers; ~1–3 min each
portfolio health, portfolio-health, 持仓健康, 持仓体检, 健康扫描	dsa_port_health.py	reads portfolio_truth.json, ~5–10 min total
market brief HK, brief HK, 市场简报 HK, 港股简报	dsa_market_brief.py HK	~30 s
market brief US, brief US, 市场简报 US, 美股简报	dsa_market_brief.py US	~30 s

OUTPUT RULE: Every script prints __REPORT_START__ ... __REPORT_END__ to stdout. The text between the markers IS the full DSA-style report (plus a trailing DONE:/FAILED: status line) — your job is to reply with that text verbatim so the user receives the report on your reply channel. Do NOT paraphrase, summarize, or add commentary.

Trigger word disambiguation

• Exact portfolio → use the existing portfolio-report skill (live P&L). NOT this skill.
• portfolio health / 持仓健康 → THIS skill (dsa_port_health.py).
• ta TICKER / research TICKER / 研究 TICKER → use the existing trading-agents-research skill (~5 min deep multi-agent analysis). NOT this skill.
• scan TICKER / quick TICKER → THIS skill (~1–3 min light technical scan).
• The two are complementary: use scan for daily monitoring, ta for high-conviction deep dives.

How to invoke

1. Single or multi ticker quick scan

User says: scan NVDA MSFT

Run via execute_code:

import subprocess, sys
result = subprocess.run(
    [
        "${PLUGIN_ROOT}/.venvs/dsa/bin/python",
        "${PLUGIN_ROOT}/skills/finance/daily-stock-analysis/scripts/dsa_scan.py",
        "NVDA", "MSFT",
    ],
    capture_output=True, text=True, timeout=900,
)
print(result.stdout)
sys.stderr.write(result.stderr)

2. Portfolio batch health scan

User says: portfolio health

import subprocess, sys
result = subprocess.run(
    [
        "${PLUGIN_ROOT}/.venvs/dsa/bin/python",
        "${PLUGIN_ROOT}/skills/finance/daily-stock-analysis/scripts/dsa_port_health.py",
    ],
    capture_output=True, text=True, timeout=1800,
)
print(result.stdout)
sys.stderr.write(result.stderr)

3. Market brief

User says: brief US

import subprocess, sys
result = subprocess.run(
    [
        "${PLUGIN_ROOT}/.venvs/dsa/bin/python",
        "${PLUGIN_ROOT}/skills/finance/daily-stock-analysis/scripts/dsa_market_brief.py",
        "US",
    ],
    capture_output=True, text=True, timeout=300,
)
print(result.stdout)
sys.stderr.write(result.stderr)

Validation

Look at the tool result. It MUST contain __REPORT_START__ and __REPORT_END__ with non-empty body between them.

• If markers exist → reply ONLY with the text between markers, verbatim. That text is the full DSA-style report the user is waiting for.
• If markers are missing → reply ONLY: ERROR: daily-stock-analysis script failed to produce output. Check stderr. Do NOT fabricate analysis.

Model routing

Profile	Primary	Fallback	Used by
fast_scan	deepseek-v4-pro via AISA gateway (always reasons)	gemini/gemini-2.5-flash (thinking off)	dsa_scan.py, dsa_port_health.py, dsa_market_brief.py

Configured inside _dsa_lib.get_router(). No environment overrides needed.

Routing decision (2026-04-30, user spec): T1/T2/T3 all use deepseek-v4-pro as primary, gemini-2.5-flash as fallback. Quality first — DeepSeek-v4-pro provides the strongest structured-output calibration and Chinese financial reasoning. Gemini is a free-tier safety net (limited to 20 requests/day) for when the AISA gateway (which serves the DeepSeek models) is unreachable.

Resources

• Scripts: ${PLUGIN_ROOT}/skills/finance/daily-stock-analysis/scripts/
• Shared lib: _dsa_lib.py (model routing, yfinance, formatting, report markers)
• Venv python: ${PLUGIN_ROOT}/.venvs/dsa/bin/python (managed by the platform's setup.python step)
• Holdings: ~/.aisa/agents/aisa-cio/portfolio/portfolio_truth.json (read-only)

Hard guarantees

• Never modifies portfolio_truth.json or any portfolio data file.
• Uses the isolated venv behind ${PLUGIN_ROOT}/.venvs/dsa/bin/python; cannot affect Hermes's own Python environment.
• Falls back to DeepSeek if Gemini is unreachable.
• Does NOT include A-share coverage; tickers like 600519/000001 will return "no data" and be skipped.