datasinking 0.2.12__tar.gz → 0.2.14__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,13 +1,13 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: datasinking
3
- Version: 0.2.12
4
- Summary: Python client for DataSinking — full-text Asian financial reports (China, Korea, Japan, Taiwan) as Markdown.
3
+ Version: 0.2.14
4
+ Summary: Python client for DataSinking — full-text financial reports (US, China, Japan, Korea, Taiwan) as Markdown.
5
5
  Author: DataSinking
6
6
  License: MIT
7
7
  Project-URL: Homepage, https://datasink.ing
8
8
  Project-URL: Documentation, https://datasink.ing/docs
9
9
  Project-URL: Repository, https://github.com/heubme2020/datasinking
10
- Keywords: a-share,china,korea,japan,financial reports,markdown,api,mcp,model-context-protocol,rag,llm
10
+ Keywords: a-share,china,korea,japan,us,sec,financial reports,markdown,api,mcp,model-context-protocol,rag,llm
11
11
  Classifier: Programming Language :: Python :: 3
12
12
  Classifier: License :: OSI Approved :: MIT License
13
13
  Classifier: Operating System :: OS Independent
@@ -28,11 +28,11 @@ Dynamic: license-file
28
28
  [![PyPI version](https://img.shields.io/pypi/v/datasinking.svg)](https://pypi.org/project/datasinking/)
29
29
  [![MCP server](https://img.shields.io/badge/MCP-server-blue)](https://github.com/heubme2020/datasinking#mcp-server)
30
30
 
31
- **Full-text financial reports across Asia, as clean Markdown.**
31
+ **Full-text financial reports across the US and Asia, as clean Markdown.**
32
32
 
33
33
  [DataSinking](https://datasink.ing) serves **full-text financial reports** — annual, semi-annual
34
- and quarterly — from **China, Korea, Japan and Taiwan** as clean **Markdown**, ready for LLM reading
35
- and RAG. Query by FMP-style symbol (`600519.SS`, `005930.KS`, `7203.T`, `2330.TW`) or filter by
34
+ and quarterly — from **the US, China, Japan, Korea and Taiwan** as clean **Markdown**, ready for LLM reading
35
+ and RAG. Query by FMP-style symbol (`AAPL`, `600519.SS`, `7203.T`, `005930.KS`, `2330.TW`) or filter by
36
36
  exchange, report period, or **section** — pull just the MD&A / risk section instead of the whole
37
37
  report. Reports are sourced from official disclosure platforms and parsed into structured Markdown
38
38
  with YAML frontmatter, preserved headings, paragraphs and tables.
@@ -41,20 +41,21 @@ with YAML frontmatter, preserved headings, paragraphs and tables.
41
41
 
42
42
  ## MCP server
43
43
 
44
- Ship DataSinking to any AI agent (Claude / Cursor / Codex / Windsurf) as an
44
+ Ship DataSinking to any AI agent (Claude / Cursor / Codex / Devin Desktop) as an
45
45
  [MCP](https://modelcontextprotocol.io) server — 6 tools: list exchanges, list stocks,
46
46
  list reports, fetch a report, list sections, fetch one section (token-friendly for RAG).
47
47
 
48
48
  ### Hosted — nothing to install
49
49
 
50
- Point any MCP client at our endpoint and you're done. No package, no Python, no local server:
50
+ One URL, no package, no Python, no local server. Most clients take this shape:
51
51
 
52
52
  ```json
53
53
  {
54
54
  "mcpServers": {
55
55
  "datasinking": {
56
56
  "type": "http",
57
- "url": "https://api.datasink.ing/mcp?apikey=YOUR_KEY"
57
+ "url": "https://api.datasink.ing/mcp",
58
+ "headers": { "Authorization": "Bearer YOUR_KEY" }
58
59
  }
59
60
  }
60
61
  }
@@ -67,9 +68,12 @@ claude mcp add --transport http datasinking https://api.datasink.ing/mcp \
67
68
  --header "Authorization: Bearer YOUR_KEY"
68
69
  ```
69
70
 
70
- Your key rides inside the URL, so treat that config as a secret. Clients that support custom
71
- headers can send `Authorization: Bearer YOUR_KEY` instead — Claude Code redacts headers in its
72
- output but can't redact a URL.
71
+ A client that can't set a header can put the key in the URL instead —
72
+ `https://api.datasink.ing/mcp?apikey=YOUR_KEY` — but only do that when you must: a key in a URL
73
+ ends up in logs and screen shares, and no client can redact it.
74
+
75
+ Two clients don't fit the shape above: Codex's config is TOML, and Devin Desktop's remote field is
76
+ `serverUrl`, not `url`. Per-client files: [`docs/mcp/`](docs/mcp/).
73
77
 
74
78
  ### Local — run it yourself
75
79
 
@@ -104,8 +108,9 @@ kept in lockstep by [`check_mcp_parity.py`](check_mcp_parity.py).
104
108
 
105
109
  Full per-client setup: [`mcp-server.md`](mcp-server.md) (overview) ·
106
110
  [`docs/mcp/`](docs/mcp/) (one guide per client).
107
- Hosted endpoint is **POST-only and stateless** — `GET /mcp` returns 405, no session id, and every
108
- client needs a restart after a config change. That trips up most first connections.
111
+
112
+ The hosted endpoint is **POST-only and stateless** — `GET /mcp` returns 405, and no session id is
113
+ issued. Most clients read MCP config at launch, so restart one after changing its config.
109
114
 
110
115
  ![DataSinking MCP in Claude](docs/images/mcp-demo.png)
111
116
 
@@ -119,8 +124,8 @@ datasinking/
119
124
  ├── research/ # Research notes / blog posts (reproducing paper-style presentation)
120
125
  ├── datasinking/ # Python client + MCP server — pip install "datasinking[mcp]"
121
126
  ├── npm/ # The same MCP server on npm — npx -y datasinking-mcp (Node 18+)
122
- ├── docs/mcp/ # Per-client MCP setup: Claude Code, Claude Desktop, Cursor, Codex, WorkBuddy
123
- ├── mcp-server.md # MCP server overview — the two config shapes, tools, endpoint limits
127
+ ├── docs/mcp/ # Per-client MCP setup — Claude Code, Desktop, Cursor, Codex, WorkBuddy, Devin…
128
+ ├── mcp-server.md # MCP server overview — endpoint, per-client table, tools, troubleshooting
124
129
  ├── llm-examples.md # Ask an LLM — no code needed (8 end-to-end examples)
125
130
  ├── api-examples.md # 7 examples × 3 interfaces (curl / Python / LLM)
126
131
  └── README.md
@@ -135,9 +140,13 @@ errors that client actually produces:
135
140
  | Claude Desktop | [`docs/mcp/claude-desktop.md`](docs/mcp/claude-desktop.md) |
136
141
  | OpenAI Codex CLI | [`docs/mcp/codex.md`](docs/mcp/codex.md) |
137
142
  | WorkBuddy / CodeBuddy | [`docs/mcp/workbuddy.md`](docs/mcp/workbuddy.md) |
143
+ | Doubao Work / 豆包工作 | [`docs/mcp/doubao.md`](docs/mcp/doubao.md) |
138
144
  | Cursor | [`docs/mcp/cursor.md`](docs/mcp/cursor.md) |
139
- | DeepSeek | [`docs/mcp/deepseek.md`](docs/mcp/deepseek.md) |
140
- | Windsurf | [`docs/mcp/windsurf.md`](docs/mcp/windsurf.md) |
145
+ | DeepSeek Harness (`dsh`) | [`docs/mcp/deepseek.md`](docs/mcp/deepseek.md) |
146
+ | Devin Desktop | [`docs/mcp/windsurf.md`](docs/mcp/windsurf.md) |
147
+ | OpenCode | [`docs/mcp/opencode.md`](docs/mcp/opencode.md) |
148
+ | Qoder | [`docs/mcp/qoder.md`](docs/mcp/qoder.md) |
149
+ | Reasonix | [`docs/mcp/reasonix.md`](docs/mcp/reasonix.md) |
141
150
 
142
151
  ## Quick start
143
152
 
@@ -197,12 +206,12 @@ Start from [`research/TEMPLATE.md`](research/TEMPLATE.md).
197
206
 
198
207
  | | |
199
208
  |---|---|
200
- | Coverage | China (SSE / SZSE / BSE) · Korea (KOSPI / KOSDAQ / KONEX) · Japan (TSE) · Taiwan (TWSE / TPEx) |
209
+ | Coverage | US (NYSE / NASDAQ / CBOE) · China (SSE / SZSE / BSE) · Japan (TSE) · Korea (KOSPI / KOSDAQ / KONEX) · Taiwan (TWSE / TPEx) |
201
210
  | Document types | annual / semiannual / q1 / q3 / amendment |
202
- | Update frequency | Daily — Korea/Japan via official DART/EDINET APIs (new filings within ~24h of publication) |
211
+ | Update frequency | Daily — US/Korea/Japan via official SEC EDGAR / DART / EDINET APIs (new filings within ~24h of publication) |
203
212
  | Format | Full-text Markdown (with YAML frontmatter) |
204
213
  | API | REST — `GET /documents`, batch download, `with_content=1` for full text, `?section=` + `/sections` for chapter-level access |
205
- | Symbols | FMP style: `600519.SS` / `005930.KS` / `7203.T` |
214
+ | Symbols | FMP style: `AAPL` / `600519.SS` / `7203.T` / `005930.KS` |
206
215
  | Auth | `?apikey=` query parameter (FMP style) |
207
216
 
208
217
  ## Data source
@@ -211,6 +220,7 @@ Reports are sourced from the official regulatory disclosure platform of each mar
211
220
 
212
221
  | Market | Source | Platform |
213
222
  |---|---|---|
223
+ | US (bare ticker, e.g. `AAPL`) | SEC EDGAR | US Securities and Exchange Commission — sec.gov/edgar |
214
224
  | China A-shares (`.SS` `.SZ` `.BJ`) | 巨潮资讯网 cninfo | CSRC-designated disclosure platform |
215
225
  | Korea (`.KS` `.KQ` `.KN`) | DART | Financial Supervisory Service — opendart.fss.or.kr |
216
226
  | Japan (`.T`) | EDINET | Financial Services Agency — disclosure2.edinet-fsa.go.jp |
@@ -5,11 +5,11 @@
5
5
  [![PyPI version](https://img.shields.io/pypi/v/datasinking.svg)](https://pypi.org/project/datasinking/)
6
6
  [![MCP server](https://img.shields.io/badge/MCP-server-blue)](https://github.com/heubme2020/datasinking#mcp-server)
7
7
 
8
- **Full-text financial reports across Asia, as clean Markdown.**
8
+ **Full-text financial reports across the US and Asia, as clean Markdown.**
9
9
 
10
10
  [DataSinking](https://datasink.ing) serves **full-text financial reports** — annual, semi-annual
11
- and quarterly — from **China, Korea, Japan and Taiwan** as clean **Markdown**, ready for LLM reading
12
- and RAG. Query by FMP-style symbol (`600519.SS`, `005930.KS`, `7203.T`, `2330.TW`) or filter by
11
+ and quarterly — from **the US, China, Japan, Korea and Taiwan** as clean **Markdown**, ready for LLM reading
12
+ and RAG. Query by FMP-style symbol (`AAPL`, `600519.SS`, `7203.T`, `005930.KS`, `2330.TW`) or filter by
13
13
  exchange, report period, or **section** — pull just the MD&A / risk section instead of the whole
14
14
  report. Reports are sourced from official disclosure platforms and parsed into structured Markdown
15
15
  with YAML frontmatter, preserved headings, paragraphs and tables.
@@ -18,20 +18,21 @@ with YAML frontmatter, preserved headings, paragraphs and tables.
18
18
 
19
19
  ## MCP server
20
20
 
21
- Ship DataSinking to any AI agent (Claude / Cursor / Codex / Windsurf) as an
21
+ Ship DataSinking to any AI agent (Claude / Cursor / Codex / Devin Desktop) as an
22
22
  [MCP](https://modelcontextprotocol.io) server — 6 tools: list exchanges, list stocks,
23
23
  list reports, fetch a report, list sections, fetch one section (token-friendly for RAG).
24
24
 
25
25
  ### Hosted — nothing to install
26
26
 
27
- Point any MCP client at our endpoint and you're done. No package, no Python, no local server:
27
+ One URL, no package, no Python, no local server. Most clients take this shape:
28
28
 
29
29
  ```json
30
30
  {
31
31
  "mcpServers": {
32
32
  "datasinking": {
33
33
  "type": "http",
34
- "url": "https://api.datasink.ing/mcp?apikey=YOUR_KEY"
34
+ "url": "https://api.datasink.ing/mcp",
35
+ "headers": { "Authorization": "Bearer YOUR_KEY" }
35
36
  }
36
37
  }
37
38
  }
@@ -44,9 +45,12 @@ claude mcp add --transport http datasinking https://api.datasink.ing/mcp \
44
45
  --header "Authorization: Bearer YOUR_KEY"
45
46
  ```
46
47
 
47
- Your key rides inside the URL, so treat that config as a secret. Clients that support custom
48
- headers can send `Authorization: Bearer YOUR_KEY` instead — Claude Code redacts headers in its
49
- output but can't redact a URL.
48
+ A client that can't set a header can put the key in the URL instead —
49
+ `https://api.datasink.ing/mcp?apikey=YOUR_KEY` — but only do that when you must: a key in a URL
50
+ ends up in logs and screen shares, and no client can redact it.
51
+
52
+ Two clients don't fit the shape above: Codex's config is TOML, and Devin Desktop's remote field is
53
+ `serverUrl`, not `url`. Per-client files: [`docs/mcp/`](docs/mcp/).
50
54
 
51
55
  ### Local — run it yourself
52
56
 
@@ -81,8 +85,9 @@ kept in lockstep by [`check_mcp_parity.py`](check_mcp_parity.py).
81
85
 
82
86
  Full per-client setup: [`mcp-server.md`](mcp-server.md) (overview) ·
83
87
  [`docs/mcp/`](docs/mcp/) (one guide per client).
84
- Hosted endpoint is **POST-only and stateless** — `GET /mcp` returns 405, no session id, and every
85
- client needs a restart after a config change. That trips up most first connections.
88
+
89
+ The hosted endpoint is **POST-only and stateless** — `GET /mcp` returns 405, and no session id is
90
+ issued. Most clients read MCP config at launch, so restart one after changing its config.
86
91
 
87
92
  ![DataSinking MCP in Claude](docs/images/mcp-demo.png)
88
93
 
@@ -96,8 +101,8 @@ datasinking/
96
101
  ├── research/ # Research notes / blog posts (reproducing paper-style presentation)
97
102
  ├── datasinking/ # Python client + MCP server — pip install "datasinking[mcp]"
98
103
  ├── npm/ # The same MCP server on npm — npx -y datasinking-mcp (Node 18+)
99
- ├── docs/mcp/ # Per-client MCP setup: Claude Code, Claude Desktop, Cursor, Codex, WorkBuddy
100
- ├── mcp-server.md # MCP server overview — the two config shapes, tools, endpoint limits
104
+ ├── docs/mcp/ # Per-client MCP setup — Claude Code, Desktop, Cursor, Codex, WorkBuddy, Devin…
105
+ ├── mcp-server.md # MCP server overview — endpoint, per-client table, tools, troubleshooting
101
106
  ├── llm-examples.md # Ask an LLM — no code needed (8 end-to-end examples)
102
107
  ├── api-examples.md # 7 examples × 3 interfaces (curl / Python / LLM)
103
108
  └── README.md
@@ -112,9 +117,13 @@ errors that client actually produces:
112
117
  | Claude Desktop | [`docs/mcp/claude-desktop.md`](docs/mcp/claude-desktop.md) |
113
118
  | OpenAI Codex CLI | [`docs/mcp/codex.md`](docs/mcp/codex.md) |
114
119
  | WorkBuddy / CodeBuddy | [`docs/mcp/workbuddy.md`](docs/mcp/workbuddy.md) |
120
+ | Doubao Work / 豆包工作 | [`docs/mcp/doubao.md`](docs/mcp/doubao.md) |
115
121
  | Cursor | [`docs/mcp/cursor.md`](docs/mcp/cursor.md) |
116
- | DeepSeek | [`docs/mcp/deepseek.md`](docs/mcp/deepseek.md) |
117
- | Windsurf | [`docs/mcp/windsurf.md`](docs/mcp/windsurf.md) |
122
+ | DeepSeek Harness (`dsh`) | [`docs/mcp/deepseek.md`](docs/mcp/deepseek.md) |
123
+ | Devin Desktop | [`docs/mcp/windsurf.md`](docs/mcp/windsurf.md) |
124
+ | OpenCode | [`docs/mcp/opencode.md`](docs/mcp/opencode.md) |
125
+ | Qoder | [`docs/mcp/qoder.md`](docs/mcp/qoder.md) |
126
+ | Reasonix | [`docs/mcp/reasonix.md`](docs/mcp/reasonix.md) |
118
127
 
119
128
  ## Quick start
120
129
 
@@ -174,12 +183,12 @@ Start from [`research/TEMPLATE.md`](research/TEMPLATE.md).
174
183
 
175
184
  | | |
176
185
  |---|---|
177
- | Coverage | China (SSE / SZSE / BSE) · Korea (KOSPI / KOSDAQ / KONEX) · Japan (TSE) · Taiwan (TWSE / TPEx) |
186
+ | Coverage | US (NYSE / NASDAQ / CBOE) · China (SSE / SZSE / BSE) · Japan (TSE) · Korea (KOSPI / KOSDAQ / KONEX) · Taiwan (TWSE / TPEx) |
178
187
  | Document types | annual / semiannual / q1 / q3 / amendment |
179
- | Update frequency | Daily — Korea/Japan via official DART/EDINET APIs (new filings within ~24h of publication) |
188
+ | Update frequency | Daily — US/Korea/Japan via official SEC EDGAR / DART / EDINET APIs (new filings within ~24h of publication) |
180
189
  | Format | Full-text Markdown (with YAML frontmatter) |
181
190
  | API | REST — `GET /documents`, batch download, `with_content=1` for full text, `?section=` + `/sections` for chapter-level access |
182
- | Symbols | FMP style: `600519.SS` / `005930.KS` / `7203.T` |
191
+ | Symbols | FMP style: `AAPL` / `600519.SS` / `7203.T` / `005930.KS` |
183
192
  | Auth | `?apikey=` query parameter (FMP style) |
184
193
 
185
194
  ## Data source
@@ -188,6 +197,7 @@ Reports are sourced from the official regulatory disclosure platform of each mar
188
197
 
189
198
  | Market | Source | Platform |
190
199
  |---|---|---|
200
+ | US (bare ticker, e.g. `AAPL`) | SEC EDGAR | US Securities and Exchange Commission — sec.gov/edgar |
191
201
  | China A-shares (`.SS` `.SZ` `.BJ`) | 巨潮资讯网 cninfo | CSRC-designated disclosure platform |
192
202
  | Korea (`.KS` `.KQ` `.KN`) | DART | Financial Supervisory Service — opendart.fss.or.kr |
193
203
  | Japan (`.T`) | EDINET | Financial Services Agency — disclosure2.edinet-fsa.go.jp |
@@ -8,4 +8,4 @@ pyproject.toml 用 `dynamic = ["version"]` 从这里读,`client.py` 的 User-A
8
8
  写着旧版本)。现在只有这一处需要改。
9
9
  """
10
10
 
11
- __version__ = "0.2.12"
11
+ __version__ = "0.2.14"
@@ -1,9 +1,9 @@
1
1
  # -*- coding: utf-8 -*-
2
2
  """DataSinking MCP server (Model Context Protocol).
3
3
 
4
- Expose the DataSinking API — full-text financial reports across Asia
5
- (China, Korea, Japan, Taiwan) as clean Markdown — to AI agents (Claude, Cursor,
6
- Codex, DeepSeek, Windsurf, …).
4
+ Expose the DataSinking API — full-text financial reports across the US and Asia
5
+ (the US, China, Japan, Korea, Taiwan) as clean Markdown — to AI agents (Claude, Cursor,
6
+ Codex, DeepSeek, Devin Desktop, …).
7
7
 
8
8
  Install the MCP extra::
9
9
 
@@ -62,14 +62,14 @@ API_KEY = os.environ.get("DATASINK_API_KEY", "")
62
62
 
63
63
  mcp = FastMCP(
64
64
  "DataSinking",
65
- title="DataSinking — Full-text Asian Financial Reports",
66
- description="Full-text Asian financial reports (China, Korea, Japan, Taiwan) as clean Markdown via API, with chapter-level access for RAG and AI agents.",
65
+ title="DataSinking — Full-text Financial Reports",
66
+ description="Full-text financial reports (US, China, Japan, Korea, Taiwan) as clean Markdown via API, with chapter-level access for RAG and AI agents.",
67
67
  version=__version__,
68
68
  instructions=(
69
69
  "DataSinking serves full-text financial reports (annual / semi-annual / quarterly) "
70
- "from China, Korea, Japan and Taiwan as clean Markdown, ready for LLM reading and RAG. "
71
- "Use FMP-style symbols: 600519.SS (Kweichow Moutai), 005930.KS (Samsung Electronics), "
72
- "7203.T (Toyota), 2330.TW (TSMC). To save tokens, prefer get_section to pull one chapter "
70
+ "from the US, China, Japan, Korea and Taiwan as clean Markdown, ready for LLM reading and RAG. "
71
+ "Use FMP-style symbols: AAPL (Apple), 600519.SS (Kweichow Moutai), 7203.T (Toyota), "
72
+ "005930.KS (Samsung Electronics), 2330.TW (TSMC). To save tokens, prefer get_section to pull one chapter "
73
73
  "(e.g. MD&A) instead of get_report for the whole document."
74
74
  ),
75
75
  )
@@ -115,9 +115,9 @@ def _get(path: str, params: Optional[dict] = None) -> dict:
115
115
  def list_exchanges() -> list:
116
116
  """List the exchanges DataSinking covers.
117
117
 
118
- Returns exchange codes (sse / szse / bj / ksc / koe / knx / jpx / twse / tpex) with
118
+ Returns exchange codes (nyse / nasdaq / cboe / sse / szse / bj / jpx / ksc / koe / knx / twse / tpex) with
119
119
  the number of reports available per exchange. Call this first to discover coverage.
120
- Sources: A-shares = cninfo.com.cn, Korea = DART, Japan = EDINET, Taiwan = MOPS.
120
+ Sources: US = SEC EDGAR, A-shares = cninfo.com.cn, Japan = EDINET, Korea = DART, Taiwan = MOPS.
121
121
  """
122
122
  return _get("/exchanges").get("exchanges", [])
123
123
 
@@ -125,7 +125,7 @@ def list_exchanges() -> list:
125
125
  @mcp.tool()
126
126
  def list_stocks(
127
127
  exchange: Annotated[
128
- str, Field(description="Exchange code, e.g. sse / szse / bj / ksc / koe / knx / jpx / twse / tpex")
128
+ str, Field(description="Exchange code, e.g. nyse / nasdaq / cboe / sse / szse / bj / jpx / ksc / koe / knx / twse / tpex")
129
129
  ],
130
130
  limit: Annotated[
131
131
  int, Field(description="Return only the first N companies (default 20) to keep the response short.")
@@ -139,7 +139,7 @@ def list_stocks(
139
139
  @mcp.tool()
140
140
  def list_reports(
141
141
  symbol: Annotated[
142
- str, Field(description="FMP-style symbol, e.g. 600519.SS / 005930.KS / 7203.T / 2330.TW")
142
+ str, Field(description="FMP-style symbol, e.g. AAPL / 600519.SS / 7203.T / 005930.KS / 2330.TW")
143
143
  ],
144
144
  doc_type: Annotated[
145
145
  str, Field(description="Report type to filter on. Defaults to annual.")
@@ -190,12 +190,13 @@ def get_section(
190
190
  str,
191
191
  Field(
192
192
  description=(
193
- "Heading keyword, matched as a substring against the report's OWN headings, "
194
- "so pass it in the report's language. A-share reports have Chinese headings "
195
- "(e.g. 第三节管理层讨论与分析) — use 管理层讨论与分析 / 财务报告 there. "
196
- "For English-language filings, \"MD&A\" / \"financial statements\" / \"notes\" work. "
197
- "If nothing matches, the API returns 404 with the real headings — retry with one "
198
- "of those, or call list_sections first."
193
+ "Heading keyword. Matches the report's own headings, and also understands "
194
+ "cross-market concepts — MD&A / financial statements / notes / audit report / "
195
+ "risk / governance (or their Chinese/Japanese/Korean equivalents) resolve to the "
196
+ "right section in any market. Otherwise pass a heading in the report's language "
197
+ "(A-share: 管理层讨论与分析 / 财务报告; English filings: \"MD&A\" / "
198
+ "\"financial statements\" / \"notes\"). If nothing matches, the API returns 404 "
199
+ "with the real headings — retry with one of those, or call list_sections first."
199
200
  )
200
201
  ),
201
202
  ],
@@ -1,13 +1,13 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: datasinking
3
- Version: 0.2.12
4
- Summary: Python client for DataSinking — full-text Asian financial reports (China, Korea, Japan, Taiwan) as Markdown.
3
+ Version: 0.2.14
4
+ Summary: Python client for DataSinking — full-text financial reports (US, China, Japan, Korea, Taiwan) as Markdown.
5
5
  Author: DataSinking
6
6
  License: MIT
7
7
  Project-URL: Homepage, https://datasink.ing
8
8
  Project-URL: Documentation, https://datasink.ing/docs
9
9
  Project-URL: Repository, https://github.com/heubme2020/datasinking
10
- Keywords: a-share,china,korea,japan,financial reports,markdown,api,mcp,model-context-protocol,rag,llm
10
+ Keywords: a-share,china,korea,japan,us,sec,financial reports,markdown,api,mcp,model-context-protocol,rag,llm
11
11
  Classifier: Programming Language :: Python :: 3
12
12
  Classifier: License :: OSI Approved :: MIT License
13
13
  Classifier: Operating System :: OS Independent
@@ -28,11 +28,11 @@ Dynamic: license-file
28
28
  [![PyPI version](https://img.shields.io/pypi/v/datasinking.svg)](https://pypi.org/project/datasinking/)
29
29
  [![MCP server](https://img.shields.io/badge/MCP-server-blue)](https://github.com/heubme2020/datasinking#mcp-server)
30
30
 
31
- **Full-text financial reports across Asia, as clean Markdown.**
31
+ **Full-text financial reports across the US and Asia, as clean Markdown.**
32
32
 
33
33
  [DataSinking](https://datasink.ing) serves **full-text financial reports** — annual, semi-annual
34
- and quarterly — from **China, Korea, Japan and Taiwan** as clean **Markdown**, ready for LLM reading
35
- and RAG. Query by FMP-style symbol (`600519.SS`, `005930.KS`, `7203.T`, `2330.TW`) or filter by
34
+ and quarterly — from **the US, China, Japan, Korea and Taiwan** as clean **Markdown**, ready for LLM reading
35
+ and RAG. Query by FMP-style symbol (`AAPL`, `600519.SS`, `7203.T`, `005930.KS`, `2330.TW`) or filter by
36
36
  exchange, report period, or **section** — pull just the MD&A / risk section instead of the whole
37
37
  report. Reports are sourced from official disclosure platforms and parsed into structured Markdown
38
38
  with YAML frontmatter, preserved headings, paragraphs and tables.
@@ -41,20 +41,21 @@ with YAML frontmatter, preserved headings, paragraphs and tables.
41
41
 
42
42
  ## MCP server
43
43
 
44
- Ship DataSinking to any AI agent (Claude / Cursor / Codex / Windsurf) as an
44
+ Ship DataSinking to any AI agent (Claude / Cursor / Codex / Devin Desktop) as an
45
45
  [MCP](https://modelcontextprotocol.io) server — 6 tools: list exchanges, list stocks,
46
46
  list reports, fetch a report, list sections, fetch one section (token-friendly for RAG).
47
47
 
48
48
  ### Hosted — nothing to install
49
49
 
50
- Point any MCP client at our endpoint and you're done. No package, no Python, no local server:
50
+ One URL, no package, no Python, no local server. Most clients take this shape:
51
51
 
52
52
  ```json
53
53
  {
54
54
  "mcpServers": {
55
55
  "datasinking": {
56
56
  "type": "http",
57
- "url": "https://api.datasink.ing/mcp?apikey=YOUR_KEY"
57
+ "url": "https://api.datasink.ing/mcp",
58
+ "headers": { "Authorization": "Bearer YOUR_KEY" }
58
59
  }
59
60
  }
60
61
  }
@@ -67,9 +68,12 @@ claude mcp add --transport http datasinking https://api.datasink.ing/mcp \
67
68
  --header "Authorization: Bearer YOUR_KEY"
68
69
  ```
69
70
 
70
- Your key rides inside the URL, so treat that config as a secret. Clients that support custom
71
- headers can send `Authorization: Bearer YOUR_KEY` instead — Claude Code redacts headers in its
72
- output but can't redact a URL.
71
+ A client that can't set a header can put the key in the URL instead —
72
+ `https://api.datasink.ing/mcp?apikey=YOUR_KEY` — but only do that when you must: a key in a URL
73
+ ends up in logs and screen shares, and no client can redact it.
74
+
75
+ Two clients don't fit the shape above: Codex's config is TOML, and Devin Desktop's remote field is
76
+ `serverUrl`, not `url`. Per-client files: [`docs/mcp/`](docs/mcp/).
73
77
 
74
78
  ### Local — run it yourself
75
79
 
@@ -104,8 +108,9 @@ kept in lockstep by [`check_mcp_parity.py`](check_mcp_parity.py).
104
108
 
105
109
  Full per-client setup: [`mcp-server.md`](mcp-server.md) (overview) ·
106
110
  [`docs/mcp/`](docs/mcp/) (one guide per client).
107
- Hosted endpoint is **POST-only and stateless** — `GET /mcp` returns 405, no session id, and every
108
- client needs a restart after a config change. That trips up most first connections.
111
+
112
+ The hosted endpoint is **POST-only and stateless** — `GET /mcp` returns 405, and no session id is
113
+ issued. Most clients read MCP config at launch, so restart one after changing its config.
109
114
 
110
115
  ![DataSinking MCP in Claude](docs/images/mcp-demo.png)
111
116
 
@@ -119,8 +124,8 @@ datasinking/
119
124
  ├── research/ # Research notes / blog posts (reproducing paper-style presentation)
120
125
  ├── datasinking/ # Python client + MCP server — pip install "datasinking[mcp]"
121
126
  ├── npm/ # The same MCP server on npm — npx -y datasinking-mcp (Node 18+)
122
- ├── docs/mcp/ # Per-client MCP setup: Claude Code, Claude Desktop, Cursor, Codex, WorkBuddy
123
- ├── mcp-server.md # MCP server overview — the two config shapes, tools, endpoint limits
127
+ ├── docs/mcp/ # Per-client MCP setup — Claude Code, Desktop, Cursor, Codex, WorkBuddy, Devin…
128
+ ├── mcp-server.md # MCP server overview — endpoint, per-client table, tools, troubleshooting
124
129
  ├── llm-examples.md # Ask an LLM — no code needed (8 end-to-end examples)
125
130
  ├── api-examples.md # 7 examples × 3 interfaces (curl / Python / LLM)
126
131
  └── README.md
@@ -135,9 +140,13 @@ errors that client actually produces:
135
140
  | Claude Desktop | [`docs/mcp/claude-desktop.md`](docs/mcp/claude-desktop.md) |
136
141
  | OpenAI Codex CLI | [`docs/mcp/codex.md`](docs/mcp/codex.md) |
137
142
  | WorkBuddy / CodeBuddy | [`docs/mcp/workbuddy.md`](docs/mcp/workbuddy.md) |
143
+ | Doubao Work / 豆包工作 | [`docs/mcp/doubao.md`](docs/mcp/doubao.md) |
138
144
  | Cursor | [`docs/mcp/cursor.md`](docs/mcp/cursor.md) |
139
- | DeepSeek | [`docs/mcp/deepseek.md`](docs/mcp/deepseek.md) |
140
- | Windsurf | [`docs/mcp/windsurf.md`](docs/mcp/windsurf.md) |
145
+ | DeepSeek Harness (`dsh`) | [`docs/mcp/deepseek.md`](docs/mcp/deepseek.md) |
146
+ | Devin Desktop | [`docs/mcp/windsurf.md`](docs/mcp/windsurf.md) |
147
+ | OpenCode | [`docs/mcp/opencode.md`](docs/mcp/opencode.md) |
148
+ | Qoder | [`docs/mcp/qoder.md`](docs/mcp/qoder.md) |
149
+ | Reasonix | [`docs/mcp/reasonix.md`](docs/mcp/reasonix.md) |
141
150
 
142
151
  ## Quick start
143
152
 
@@ -197,12 +206,12 @@ Start from [`research/TEMPLATE.md`](research/TEMPLATE.md).
197
206
 
198
207
  | | |
199
208
  |---|---|
200
- | Coverage | China (SSE / SZSE / BSE) · Korea (KOSPI / KOSDAQ / KONEX) · Japan (TSE) · Taiwan (TWSE / TPEx) |
209
+ | Coverage | US (NYSE / NASDAQ / CBOE) · China (SSE / SZSE / BSE) · Japan (TSE) · Korea (KOSPI / KOSDAQ / KONEX) · Taiwan (TWSE / TPEx) |
201
210
  | Document types | annual / semiannual / q1 / q3 / amendment |
202
- | Update frequency | Daily — Korea/Japan via official DART/EDINET APIs (new filings within ~24h of publication) |
211
+ | Update frequency | Daily — US/Korea/Japan via official SEC EDGAR / DART / EDINET APIs (new filings within ~24h of publication) |
203
212
  | Format | Full-text Markdown (with YAML frontmatter) |
204
213
  | API | REST — `GET /documents`, batch download, `with_content=1` for full text, `?section=` + `/sections` for chapter-level access |
205
- | Symbols | FMP style: `600519.SS` / `005930.KS` / `7203.T` |
214
+ | Symbols | FMP style: `AAPL` / `600519.SS` / `7203.T` / `005930.KS` |
206
215
  | Auth | `?apikey=` query parameter (FMP style) |
207
216
 
208
217
  ## Data source
@@ -211,6 +220,7 @@ Reports are sourced from the official regulatory disclosure platform of each mar
211
220
 
212
221
  | Market | Source | Platform |
213
222
  |---|---|---|
223
+ | US (bare ticker, e.g. `AAPL`) | SEC EDGAR | US Securities and Exchange Commission — sec.gov/edgar |
214
224
  | China A-shares (`.SS` `.SZ` `.BJ`) | 巨潮资讯网 cninfo | CSRC-designated disclosure platform |
215
225
  | Korea (`.KS` `.KQ` `.KN`) | DART | Financial Supervisory Service — opendart.fss.or.kr |
216
226
  | Japan (`.T`) | EDINET | Financial Services Agency — disclosure2.edinet-fsa.go.jp |
@@ -7,12 +7,12 @@ name = "datasinking"
7
7
  # 版本号唯一来源 = datasinking/_version.py(client.py 的 User-Agent 也读它)。
8
8
  # 以前这里和包内 __version__ 各写各的,结果漂了半年(发 0.2.4 时才发现 __init__ 还停在 0.1.0)。
9
9
  dynamic = ["version"]
10
- description = "Python client for DataSinking — full-text Asian financial reports (China, Korea, Japan, Taiwan) as Markdown."
10
+ description = "Python client for DataSinking — full-text financial reports (US, China, Japan, Korea, Taiwan) as Markdown."
11
11
  readme = "README.md"
12
12
  requires-python = ">=3.8"
13
13
  license = { text = "MIT" }
14
14
  authors = [{ name = "DataSinking" }]
15
- keywords = ["a-share", "china", "korea", "japan", "financial reports", "markdown", "api", "mcp", "model-context-protocol", "rag", "llm"]
15
+ keywords = ["a-share", "china", "korea", "japan", "us", "sec", "financial reports", "markdown", "api", "mcp", "model-context-protocol", "rag", "llm"]
16
16
  classifiers = [
17
17
  "Programming Language :: Python :: 3",
18
18
  "License :: OSI Approved :: MIT License",
File without changes
File without changes