@anysiteio/agent-skills 1.3.0 → 1.4.0

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.
@@ -132,6 +132,18 @@
132
132
  "source": "./",
133
133
  "category": "productivity",
134
134
  "skills": ["./skills/anysite-cli"]
135
+ },
136
+ {
137
+ "name": "anysite-mcp-migration",
138
+ "description": "Migration assistant for updating anysite MCP skills, prompts, and agent instructions from v1 (individual tools) to v2 (universal meta-tools: execute, discover, get_page, query_cache, export_data). Automatically rewrites tool references, adds pagination/filtering/export capabilities, and validates migrated output.",
139
+ "version": "1.0.0",
140
+ "author": {
141
+ "name": "Anysite Skills Contributors",
142
+ "email": "support@anysite.io"
143
+ },
144
+ "source": "./",
145
+ "category": "productivity",
146
+ "skills": ["./skills/anysite-mcp-migration"]
135
147
  }
136
148
  ]
137
149
  }
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@anysiteio/agent-skills",
3
- "version": "1.3.0",
3
+ "version": "1.4.0",
4
4
  "description": "Official anysite agent skills for LinkedIn intelligence, social media analysis, data extraction, and security auditing",
5
5
  "keywords": [
6
6
  "mcp",
@@ -0,0 +1,279 @@
1
+ ---
2
+ name: anysite-mcp-migration
3
+ description: Migrate anysite MCP skills, prompts, and agent instructions from v1 (individual tools like search_linkedin_users, get_linkedin_profile) to v2 (universal meta-tools execute, discover, get_page, query_cache, export_data). Automatically rewrites tool references, adds pagination/filtering/export capabilities, and validates migrated output. Use when users need to update existing skills or prompts for the new anysite MCP v2 API, convert old tool calls to execute() format, or adapt workflows to use new v2 features like server-side filtering and data export.
4
+ ---
5
+
6
+ # anysite MCP Migration Assistant
7
+
8
+ Migrate your anysite MCP skills, prompts, and agent instructions from v1 (individual tools) to v2 (universal meta-tools).
9
+
10
+ ## Overview
11
+
12
+ Anysite MCP v2 replaces 70+ individual tools with 5 universal meta-tools. This skill helps you:
13
+ - **Rewrite tool references** from old format (`search_linkedin_users`, `get_linkedin_profile`) to new `execute()` calls
14
+ - **Add new v2 capabilities** like pagination, server-side filtering, aggregation, and file export
15
+ - **Validate migrated output** to ensure nothing was missed
16
+ - **Preserve workflow logic** while updating only the tool interface layer
17
+
18
+ ## When to Use
19
+
20
+ - User says "migrate my skill to v2", "update for new anysite API", "convert to execute format"
21
+ - User pastes a skill or prompt that contains old-style tool names (`search_*`, `get_*`, `find_*`)
22
+ - User asks how to use new anysite MCP features (pagination, cache queries, export)
23
+
24
+ ---
25
+
26
+ ## Migration Workflow
27
+
28
+ ### Step 1: Receive Input
29
+
30
+ Ask the user for one of:
31
+ 1. **A skill file path** — read the SKILL.md and any files in `references/`
32
+ 2. **Pasted prompt text** — the raw text of their prompt or instruction
33
+ 3. **A description** of what the skill does — you'll help build it from scratch using v2 format
34
+
35
+ ### Step 2: Identify Old Tool References
36
+
37
+ Scan the input for any of these v1 tool name patterns:
38
+
39
+ | Pattern | Example |
40
+ |---------|---------|
41
+ | `search_linkedin_*` | `search_linkedin_users`, `search_linkedin_companies`, `search_linkedin_jobs`, `search_linkedin_posts` |
42
+ | `get_linkedin_*` | `get_linkedin_profile`, `get_linkedin_company`, `get_linkedin_user_posts` |
43
+ | `find_linkedin_*` | `find_linkedin_email`, `find_linkedin_user_email` |
44
+ | `google_linkedin_*` | `google_linkedin_search` |
45
+ | `search_twitter_*` / `get_twitter_*` | `search_twitter_users`, `get_twitter_user`, `get_twitter_user_tweets` |
46
+ | `search_instagram_*` / `get_instagram_*` | `search_instagram_users`, `get_instagram_user`, `get_instagram_post` |
47
+ | `search_youtube_*` / `get_youtube_*` | `search_youtube`, `get_youtube_channel`, `get_youtube_video` |
48
+ | `search_reddit_*` / `get_reddit_*` | `search_reddit`, `get_reddit_user`, `get_reddit_posts` |
49
+ | `search_yc_*` / `get_yc_*` | `search_yc_companies`, `get_yc_company` |
50
+ | `search_sec_*` / `get_sec_*` | `search_sec_filings`, `get_sec_document` |
51
+ | `scrape_webpage` | `scrape_webpage` |
52
+ | `mcp__anysite__*` | Any tool with the MCP prefix — strip prefix and match above |
53
+
54
+ Also look for references to **Crunchbase** — this source is disabled in v2 and must be removed.
55
+
56
+ ### Step 3: Apply Tool Mapping
57
+
58
+ Replace each old tool call using this mapping:
59
+
60
+ #### LinkedIn
61
+
62
+ | Old tool | New execute() call |
63
+ |----------|-------------------|
64
+ | `search_linkedin_users(keywords, location, count, ...)` | `execute("linkedin", "search", "search_users", {"keywords": ..., "location": ..., "count": ...})` |
65
+ | `get_linkedin_profile(user)` | `execute("linkedin", "user", "get", {"user": ...})` |
66
+ | `get_linkedin_company(company)` | `execute("linkedin", "company", "get", {"company": ...})` |
67
+ | `search_linkedin_companies(keywords, count)` | `execute("linkedin", "search", "search_companies", {"keywords": ..., "count": ...})` |
68
+ | `search_linkedin_jobs(keywords, location, count)` | `execute("linkedin", "job_search", "search_jobs", {"keywords": ..., "count": ...})` |
69
+ | `search_linkedin_posts(keywords, count)` | `execute("linkedin", "post", "search_posts", {"keywords": ..., "count": ...})` |
70
+ | `get_linkedin_user_posts(user)` | `execute("linkedin", "post", "get_user_posts", {"user": ...})` |
71
+ | `find_linkedin_email(user)` | `execute("linkedin", "email", "find", {"user": ...})` |
72
+ | `google_linkedin_search(query, count)` | `execute("linkedin", "google", "search", {"query": ..., "count": ...})` |
73
+
74
+ #### Twitter/X
75
+
76
+ | Old tool | New execute() call |
77
+ |----------|-------------------|
78
+ | `search_twitter_users(query)` | `execute("twitter", "search", "search_users", {"query": ...})` |
79
+ | `get_twitter_user(username)` | `execute("twitter", "user", "get", {"username": ...})` |
80
+ | `get_twitter_user_tweets(username)` | `execute("twitter", "user_tweets", "get", {"username": ...})` |
81
+
82
+ #### Instagram
83
+
84
+ | Old tool | New execute() call |
85
+ |----------|-------------------|
86
+ | `search_instagram_users(query)` | `execute("instagram", "search", "search_users", {"query": ...})` |
87
+ | `get_instagram_user(username)` | `execute("instagram", "user", "get", {"username": ...})` |
88
+ | `get_instagram_post(url)` | `execute("instagram", "post", "get", {"url": ...})` |
89
+
90
+ #### YouTube
91
+
92
+ | Old tool | New execute() call |
93
+ |----------|-------------------|
94
+ | `search_youtube(query, count)` | `execute("youtube", "search", "search_videos", {"query": ..., "count": ...})` |
95
+ | `get_youtube_channel(channel_id)` | `execute("youtube", "channel", "get", {"channel_id": ...})` |
96
+ | `get_youtube_video(video_id)` | `execute("youtube", "video", "get", {"video_id": ...})` |
97
+
98
+ #### Reddit
99
+
100
+ | Old tool | New execute() call |
101
+ |----------|-------------------|
102
+ | `search_reddit(query)` | `execute("reddit", "search", "search", {"query": ...})` |
103
+ | `get_reddit_user(username)` | `execute("reddit", "user", "get", {"username": ...})` |
104
+ | `get_reddit_posts(subreddit)` | `execute("reddit", "posts", "get", {"subreddit": ...})` |
105
+
106
+ #### YC / SEC / Web
107
+
108
+ | Old tool | New execute() call |
109
+ |----------|-------------------|
110
+ | `search_yc_companies(query)` | `execute("yc", "search", "search", {"query": ...})` |
111
+ | `get_yc_company(slug)` | `execute("yc", "company", "get", {"slug": ...})` |
112
+ | `search_sec_filings(query)` | `execute("sec", "search", "search", {"query": ...})` |
113
+ | `get_sec_document(url)` | `execute("sec", "document", "get", {"url": ...})` |
114
+ | `scrape_webpage(url)` | `execute("webparser", "parse", "parse", {"url": ...})` |
115
+
116
+ ### Step 4: Handle Unknown Endpoints
117
+
118
+ If the input references a tool name **not in the mapping above**:
119
+
120
+ 1. Try to infer the source and category from the tool name
121
+ 2. Add a `discover()` call before the `execute()`:
122
+ ```
123
+ Use discover("{source}", "{category}") to find available endpoints and params.
124
+ Then use execute("{source}", "{category}", "{endpoint}", {params}).
125
+ ```
126
+
127
+ **Important:** Only add `discover()` when the exact endpoint or params are unknown. If the mapping above covers it, use `execute()` directly — no discover needed.
128
+
129
+ ### Step 5: Add v2 Capabilities
130
+
131
+ Review the migrated skill for opportunities to add new v2 features:
132
+
133
+ #### Pagination
134
+ When the workflow processes large result sets or needs "more results":
135
+ ```
136
+ Results from execute() include cache_key. If more data exists, use:
137
+ get_page(cache_key="{cache_key}", offset=10, limit=10)
138
+ ```
139
+
140
+ #### Server-side Filtering
141
+ When the workflow filters results after fetching (e.g., "only show people in SF"):
142
+ ```
143
+ After execute(), filter without consuming context tokens:
144
+ query_cache(cache_key="{cache_key}", conditions=[{"field": "location", "op": "contains", "value": "San Francisco"}])
145
+ ```
146
+
147
+ #### Aggregation
148
+ When the workflow computes statistics or groups data:
149
+ ```
150
+ query_cache(cache_key="{cache_key}", aggregate={"field": "followers", "op": "avg"}, group_by="industry")
151
+ ```
152
+
153
+ #### Export to File
154
+ When the workflow outputs structured data (CSV, JSON) or the user needs downloadable results:
155
+ ```
156
+ export_data(cache_key="{cache_key}", output_format="csv")
157
+ → returns download URL
158
+ ```
159
+
160
+ ### Step 6: Update Error Handling
161
+
162
+ Replace old-style error handling:
163
+
164
+ **Before:**
165
+ ```
166
+ If search_linkedin_users returns an error, try with different keywords.
167
+ ```
168
+
169
+ **After:**
170
+ ```
171
+ If execute() returns an error with "llm_hint", follow the hint.
172
+ If execute() returns {"error": "Source not found", "available_sources": [...]}, check source name.
173
+ If execute() returns {"error": "Endpoint not found", "available_endpoints": [...]}, call discover() to find correct endpoint names.
174
+ ```
175
+
176
+ ### Step 7: Remove Disabled Sources
177
+
178
+ Remove any references to **Crunchbase** — this source is disabled in v2. If the skill relies on Crunchbase data, suggest alternatives:
179
+ - Company data → LinkedIn company profiles or Y Combinator
180
+ - Funding data → SEC filings or web scraping of funding databases
181
+ - Startup research → Y Combinator database
182
+
183
+ ### Step 8: Validate
184
+
185
+ Run through this checklist on the migrated output:
186
+
187
+ - [ ] No remaining `search_linkedin_*`, `get_linkedin_*`, `find_linkedin_*` references
188
+ - [ ] No remaining `search_twitter_*`, `get_twitter_*` references
189
+ - [ ] No remaining `search_instagram_*`, `get_instagram_*` references
190
+ - [ ] No remaining `search_youtube_*`, `get_youtube_*` references
191
+ - [ ] No remaining `search_reddit_*`, `get_reddit_*` references
192
+ - [ ] No remaining `search_yc_*`, `get_yc_*` references
193
+ - [ ] No remaining `search_sec_*`, `get_sec_*` references
194
+ - [ ] No remaining `scrape_webpage` references
195
+ - [ ] No remaining `mcp__anysite__` prefixed tool names (old MCP format)
196
+ - [ ] No Crunchbase references
197
+ - [ ] `discover()` added only where endpoint/params are genuinely unknown
198
+ - [ ] `get_page` / `query_cache` / `export_data` added where beneficial
199
+ - [ ] Error handling updated to v2 format
200
+ - [ ] Original workflow logic preserved — same steps, same data flow
201
+
202
+ ### Step 9: Output
203
+
204
+ Present the migrated skill in the same format as the input:
205
+ - If the input was a SKILL.md file → output the full migrated SKILL.md
206
+ - If the input was a prompt → output the migrated prompt text
207
+ - If the input was pasted text → output the migrated version
208
+
209
+ Always show a **migration summary** after the output:
210
+
211
+ ```
212
+ ## Migration Summary
213
+ - Tool calls replaced: N
214
+ - discover() calls added: N
215
+ - New v2 features added: [list]
216
+ - Crunchbase references removed: N
217
+ - Warnings: [any issues found]
218
+ ```
219
+
220
+ ---
221
+
222
+ ## Examples
223
+
224
+ ### Example 1: Simple Tool Replacement
225
+
226
+ **Input:**
227
+ ```
228
+ Use search_linkedin_users to find CTOs in Berlin, then get_linkedin_profile for each result.
229
+ ```
230
+
231
+ **Output:**
232
+ ```
233
+ Use execute("linkedin", "search", "search_users", {"title": "CTO", "location": "Berlin", "count": 10}) to find CTOs in Berlin, then execute("linkedin", "user", "get", {"user": "{alias}"}) for each result.
234
+ ```
235
+
236
+ ### Example 2: Multi-Step Workflow
237
+
238
+ **Input:**
239
+ ```
240
+ 1. Use search_linkedin_users to find the person
241
+ 2. Use get_linkedin_profile to get their full profile
242
+ 3. Use find_linkedin_email to get their email
243
+ 4. Use get_twitter_user to check their Twitter
244
+ ```
245
+
246
+ **Output:**
247
+ ```
248
+ 1. Use execute("linkedin", "search", "search_users", {"first_name": ..., "last_name": ..., "count": 5}) to find the person
249
+ 2. Use execute("linkedin", "user", "get", {"user": "{alias from step 1}"}) to get their full profile
250
+ 3. Use execute("linkedin", "email", "find", {"user": "{alias from step 1}"}) to get their email
251
+ 4. Use execute("twitter", "user", "get", {"username": "..."}) to check their Twitter
252
+ ```
253
+
254
+ ### Example 3: Adding v2 Features
255
+
256
+ **Input:**
257
+ ```
258
+ Search for 50 marketing managers and filter by location.
259
+ ```
260
+
261
+ **Output:**
262
+ ```
263
+ 1. Use execute("linkedin", "search", "search_users", {"title": "Marketing Manager", "count": 50}) to search
264
+ 2. If more results exist, use get_page(cache_key="{cache_key}", offset=10, limit=10) to load additional pages
265
+ 3. Use query_cache(cache_key="{cache_key}", conditions=[{"field": "location", "op": "contains", "value": "..."}]) to filter by location server-side
266
+ 4. Use export_data(cache_key="{cache_key}", output_format="csv") to download the filtered list
267
+ ```
268
+
269
+ ---
270
+
271
+ ## Quick Reference: v2 Meta-Tools
272
+
273
+ | Tool | Purpose | When to use |
274
+ |------|---------|-------------|
275
+ | `discover(source, category)` | Learn available endpoints and params | Before execute() when endpoint name or params are unknown |
276
+ | `execute(source, category, endpoint, params)` | Fetch data from any source | Every data retrieval — replaces all v1 tools |
277
+ | `get_page(cache_key, offset, limit)` | Load more items from previous execute() | When execute() returned next_offset |
278
+ | `query_cache(cache_key, conditions, sort_by, aggregate, group_by)` | Filter/sort/aggregate cached data | When you need to slice results without re-fetching |
279
+ | `export_data(cache_key, format)` | Save dataset as downloadable file | When user needs CSV/JSON/JSONL export |