context-mode 0.5.24 → 0.5.26

package/.claude-plugin/hooks/hooks.json

@@ -0,0 +1,16 @@
+{
+  "description": "Context-mode tool routing — injects MCP tool instructions into subagent prompts",
+  "hooks": {
+    "PreToolUse": [
+      {
+        "matcher": "Task",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "bash ${CLAUDE_PLUGIN_ROOT}/hooks/task-inject.sh"
+          }
+        ]
+      }
+    ]
+  }
+}

package/.claude-plugin/marketplace.json

@@ -13,7 +13,7 @@
       "name": "context-mode",
       "source": "./",
       "description": "Claude Code MCP plugin that saves 98% of your context window. Sandboxed code execution in 10 languages, FTS5 knowledge base with BM25 ranking, and intent-driven search.",
-      "version": "0.5.24",
+      "version": "0.5.26",
       "author": {
         "name": "Mert Koseoğlu"
       },

package/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "context-mode",
-  "version": "0.5.24",
+  "version": "0.5.26",
   "description": "Claude Code MCP plugin that saves 98% of your context window. Sandboxed code execution in 10 languages, FTS5 knowledge base with BM25 ranking, and intent-driven search.",
   "author": {
     "name": "Mert Koseoğlu",
@@ -25,27 +25,5 @@
       "args": ["${CLAUDE_PLUGIN_ROOT}/start.sh"]
     }
   },
-  "skills": "./skills/",
-  "hooks": {
-    "PreToolUse": [
-      {
-        "matcher": "Bash",
-        "hooks": [
-          {
-            "type": "prompt",
-            "prompt": "STOP. Default to mcp__context-mode__execute for ALL Bash commands. ONLY use Bash for these guaranteed-small-output operations: file mutations (mkdir, mv, cp, rm, touch, chmod), git writes (git add, git commit, git push, git checkout, git branch, git merge), navigation (cd, pwd, which), process control (kill, pkill), package management (npm install/publish, pip install), and simple echo/printf. Everything else — any command that reads, queries, fetches, lists, logs, tests, builds, diffs, inspects, or calls an external service — MUST use mcp__context-mode__execute instead. When uncertain, use context-mode."
-          }
-        ]
-      },
-      {
-        "matcher": "Task",
-        "hooks": [
-          {
-            "type": "prompt",
-            "prompt": "IMPORTANT: The subagent inherits MCP tools but NOT skill instructions. You MUST include these tool routing rules in the subagent prompt: 'Use mcp__context-mode__execute (language, code, intent) instead of Bash for ANY command that produces more than a few lines of output. Use mcp__context-mode__fetch_and_index (url, source) instead of WebFetch to fetch web pages. Use mcp__context-mode__search (query, source) to search indexed content. Use mcp__context-mode__execute_file (path, language, code) instead of Read for large file analysis. Do NOT use Bash, cat, curl, Read, or WebFetch for large outputs.'"
-          }
-        ]
-      }
-    ]
-  }
+  "skills": "./skills/"
 }

package/build/server.js

@@ -5,7 +5,7 @@ import { z } from "zod";
 import { PolyglotExecutor } from "./executor.js";
 import { ContentStore } from "./store.js";
 import { detectRuntimes, getRuntimeSummary, getAvailableLanguages, hasBunRuntime, } from "./runtime.js";
-const VERSION = "0.5.24";
+const VERSION = "0.5.26";
 const runtimes = detectRuntimes();
 const available = getAvailableLanguages(runtimes);
 const server = new McpServer({
@@ -30,7 +30,7 @@ const bunNote = hasBunRuntime()
 // ─────────────────────────────────────────────────────────
 server.registerTool("execute", {
     title: "Execute Code",
-    description: `Execute code in a sandboxed subprocess. Only stdout enters context — raw data stays in the subprocess. Use instead of bash/cat when output would exceed 20 lines.${bunNote} Available: ${langList}.`,
+    description: `Execute code in a sandboxed subprocess. Only stdout enters context — raw data stays in the subprocess. Use instead of bash/cat when output would exceed 20 lines.${bunNote} Available: ${langList}.\n\nPREFER THIS OVER BASH for: API calls (gh, curl, aws), test runners (npm test, pytest), git queries (git log, git diff), data processing, and ANY CLI command that may produce large output. Bash should only be used for file mutations, git writes, and navigation.`,
     inputSchema: z.object({
         language: z
             .enum([
@@ -217,7 +217,7 @@ function intentSearch(stdout, intent, source, maxResults = 5) {
 // ─────────────────────────────────────────────────────────
 server.registerTool("execute_file", {
     title: "Execute File Processing",
-    description: "Read a file and process it without loading contents into context. The file is read into a FILE_CONTENT variable inside the sandbox. Only your printed summary enters context.",
+    description: "Read a file and process it without loading contents into context. The file is read into a FILE_CONTENT variable inside the sandbox. Only your printed summary enters context.\n\nPREFER THIS OVER Read/cat for: log files, data files (CSV, JSON, XML), large source files for analysis, and any file where you need to extract specific information rather than read the entire content.",
     inputSchema: z.object({
         path: z
             .string()
@@ -521,7 +521,8 @@ server.registerTool("fetch_and_index", {
     title: "Fetch & Index URL",
     description: "Fetches URL content, converts HTML to markdown, and indexes into the searchable knowledge base. " +
         "Raw content never enters context — only a brief confirmation is returned.\n\n" +
-        "Use INSTEAD of WebFetch/Context7 when you need to reference web documentation later via search.\n\n" +
+        "PREFER THIS OVER WebFetch when you need to reference web documentation later via search. " +
+        "WebFetch loads entire page content into context; this tool indexes it and lets you search() on-demand.\n\n" +
         "After fetching, use 'search' to retrieve specific sections on-demand.",
     inputSchema: z.object({
         url: z.string().describe("The URL to fetch and index"),

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "context-mode",
-  "version": "0.5.24",
+  "version": "0.5.26",
   "type": "module",
   "description": "Claude Code MCP plugin that saves 98% of your context window. Sandboxed code execution, FTS5 knowledge base, and intent-driven search.",
   "author": "Mert Koseoğlu",

package/server.bundle.mjs

@@ -132,7 +132,9 @@ ${n}`}}};import{createRequire as aT}from"node:module";import{readFileSync as cT}
 `).trim();u.length!==0&&(r.push({title:this.#o(o,i),content:u,hasCode:s.some(l=>/^`{3,}/.test(l))}),s=[])},c=0;for(;c<n.length;){let u=n[c];if(/^[-_*]{3,}\s*$/.test(u)){a(),c++;continue}let l=u.match(/^(#{1,4})\s+(.+)$/);if(l){a();let m=l[1].length,f=l[2].trim();for(;o.length>0&&o[o.length-1].level>=m;)o.pop();o.push({level:m,text:f}),i=f,s.push(u),c++;continue}let d=u.match(/^(`{3,})(.*)?$/);if(d){let m=d[1],f=[u];for(c++;c<n.length;){if(f.push(n[c]),n[c].startsWith(m)&&n[c].trim()===m){c++;break}c++}s.push(...f);continue}s.push(u),c++}return a(),r}#n(e,r){let n=e.split(/\n\s*\n/);if(n.length>=3&&n.length<=200&&n.every(c=>Buffer.byteLength(c)<5e3))return n.map((c,u)=>{let l=c.trim();return{title:l.split(`
 `)[0].slice(0,80)||`Section ${u+1}`,content:l}}).filter(c=>c.content.length>0);let o=e.split(`
 `);if(o.length<=r)return[{title:"Output",content:e}];let s=[],a=Math.max(r-2,1);for(let c=0;c<o.length;c+=a){let u=o.slice(c,c+r);if(u.length===0)break;let l=c+1,d=Math.min(c+u.length,o.length),m=u[0]?.trim().slice(0,80);s.push({title:m||`Lines ${l}-${d}`,content:u.join(`
-`)})}return s}#o(e,r){return e.length===0?r||"Untitled":e.map(n=>n.text).join(" > ")}};var Qg="0.5.23",Wl=ai(),mT=Jg(Wl),dn=new oi({name:"context-mode",version:Qg}),Xl=new ci({runtimes:Wl}),Jl=null;function di(){return Jl||(Jl=new _o),Jl}var hT=mT.join(", "),gT=Kl()?" (Bun detected \u2014 JS/TS runs 3-5x faster)":"";dn.registerTool("execute",{title:"Execute Code",description:`Execute code in a sandboxed subprocess. Only stdout enters context \u2014 raw data stays in the subprocess. Use instead of bash/cat when output would exceed 20 lines.${gT} Available: ${hT}.`,inputSchema:ie.object({language:ie.enum(["javascript","typescript","python","shell","ruby","go","rust","php","perl","r"]).describe("Runtime language"),code:ie.string().describe("Source code to execute. Use console.log (JS/TS), print (Python/Ruby/Perl/R), echo (Shell), echo (PHP), or fmt.Println (Go) to output a summary to context."),timeout:ie.number().optional().default(3e4).describe("Max execution time in ms"),intent:ie.string().optional().describe(`What you're looking for in the output. When provided and output is large (>5KB), indexes output into knowledge base and returns section titles + previews \u2014 not full content. Use search() to retrieve specific sections. Example: 'failing tests', 'HTTP 500 errors'.
+`)})}return s}#o(e,r){return e.length===0?r||"Untitled":e.map(n=>n.text).join(" > ")}};var Qg="0.5.26",Wl=ai(),mT=Jg(Wl),dn=new oi({name:"context-mode",version:Qg}),Xl=new ci({runtimes:Wl}),Jl=null;function di(){return Jl||(Jl=new _o),Jl}var hT=mT.join(", "),gT=Kl()?" (Bun detected \u2014 JS/TS runs 3-5x faster)":"";dn.registerTool("execute",{title:"Execute Code",description:`Execute code in a sandboxed subprocess. Only stdout enters context \u2014 raw data stays in the subprocess. Use instead of bash/cat when output would exceed 20 lines.${gT} Available: ${hT}.
+
+PREFER THIS OVER BASH for: API calls (gh, curl, aws), test runners (npm test, pytest), git queries (git log, git diff), data processing, and ANY CLI command that may produce large output. Bash should only be used for file mutations, git writes, and navigation.`,inputSchema:ie.object({language:ie.enum(["javascript","typescript","python","shell","ruby","go","rust","php","perl","r"]).describe("Runtime language"),code:ie.string().describe("Source code to execute. Use console.log (JS/TS), print (Python/Ruby/Perl/R), echo (Shell), echo (PHP), or fmt.Println (Go) to output a summary to context."),timeout:ie.number().optional().default(3e4).describe("Max execution time in ms"),intent:ie.string().optional().describe(`What you're looking for in the output. When provided and output is large (>5KB), indexes output into knowledge base and returns section titles + previews \u2014 not full content. Use search() to retrieve specific sections. Example: 'failing tests', 'HTTP 500 errors'.
 
 TIP: Use specific technical terms, not just concepts. Check 'Searchable terms' in the response for available vocabulary.`)})},async({language:t,code:e,timeout:r,intent:n})=>{try{let o=await Xl.execute({language:t,code:e,timeout:r});if(o.timedOut)return{content:[{type:"text",text:`Execution timed out after ${r}ms
 
@@ -151,7 +153,9 @@ Use search() to query this content. Use source: "${n.label}" to scope results.`}
 `).length,s=Buffer.byteLength(t),i=di(),a=i.indexPlainText(t,r),c=new _o(":memory:");try{c.indexPlainText(t,r);let u=c.search(e,n);if(u.length===0){let m=e.trim().split(/\s+/).filter(f=>f.length>2).slice(0,20);if(m.length>0){let f=new Map;for(let p of m){let h=c.search(p,10);for(let g of h){let v=f.get(g.title);v?(v.score+=1,g.rank<v.bestRank&&(v.bestRank=g.rank,v.result=g)):f.set(g.title,{result:g,score:1,bestRank:g.rank})}}u=Array.from(f.values()).sort((p,h)=>h.score-p.score||p.bestRank-h.bestRank).slice(0,n).map(p=>p.result)}}let l=i.getDistinctiveTerms(a.sourceId);if(u.length===0){let m=[`Indexed ${a.totalChunks} sections from "${r}" into knowledge base.`,`No sections matched intent "${e}" in ${o}-line output (${(s/1024).toFixed(1)}KB).`];return l.length>0&&(m.push(""),m.push(`Searchable terms: ${l.join(", ")}`)),m.push(""),m.push("Use search() to explore the indexed content."),m.join(`
 `)}let d=[`Indexed ${a.totalChunks} sections from "${r}" into knowledge base.`,`${u.length} sections matched "${e}" (${o} lines, ${(s/1024).toFixed(1)}KB):`,""];for(let m of u){let f=m.content.split(`
 `)[0].slice(0,120);d.push(`  - ${m.title}: ${f}`)}return l.length>0&&(d.push(""),d.push(`Searchable terms: ${l.join(", ")}`)),d.push(""),d.push("Use search() to retrieve full content of any section."),d.join(`
-`)}finally{c.close()}}dn.registerTool("execute_file",{title:"Execute File Processing",description:"Read a file and process it without loading contents into context. The file is read into a FILE_CONTENT variable inside the sandbox. Only your printed summary enters context.",inputSchema:ie.object({path:ie.string().describe("Absolute file path or relative to project root"),language:ie.enum(["javascript","typescript","python","shell","ruby","go","rust","php","perl","r"]).describe("Runtime language"),code:ie.string().describe("Code to process FILE_CONTENT. Print summary via console.log/print/echo."),timeout:ie.number().optional().default(3e4).describe("Max execution time in ms"),intent:ie.string().optional().describe("What you're looking for in the output. When provided and output is large (>5KB), returns only matching sections via BM25 search instead of truncated output.")})},async({path:t,language:e,code:r,timeout:n,intent:o})=>{try{let s=await Xl.executeFile({path:t,language:e,code:r,timeout:n});if(s.timedOut)return{content:[{type:"text",text:`Timed out processing ${t} after ${n}ms`}],isError:!0};if(s.exitCode!==0){let a=`Error processing ${t} (exit ${s.exitCode}):
+`)}finally{c.close()}}dn.registerTool("execute_file",{title:"Execute File Processing",description:`Read a file and process it without loading contents into context. The file is read into a FILE_CONTENT variable inside the sandbox. Only your printed summary enters context.
+
+PREFER THIS OVER Read/cat for: log files, data files (CSV, JSON, XML), large source files for analysis, and any file where you need to extract specific information rather than read the entire content.`,inputSchema:ie.object({path:ie.string().describe("Absolute file path or relative to project root"),language:ie.enum(["javascript","typescript","python","shell","ruby","go","rust","php","perl","r"]).describe("Runtime language"),code:ie.string().describe("Code to process FILE_CONTENT. Print summary via console.log/print/echo."),timeout:ie.number().optional().default(3e4).describe("Max execution time in ms"),intent:ie.string().optional().describe("What you're looking for in the output. When provided and output is large (>5KB), returns only matching sections via BM25 search instead of truncated output.")})},async({path:t,language:e,code:r,timeout:n,intent:o})=>{try{let s=await Xl.executeFile({path:t,language:e,code:r,timeout:n});if(s.timedOut)return{content:[{type:"text",text:`Timed out processing ${t} after ${n}ms`}],isError:!0};if(s.exitCode!==0){let a=`Error processing ${t} (exit ${s.exitCode}):
 ${s.stderr||s.stdout}`;return o&&o.trim().length>0&&Buffer.byteLength(a)>ui?{content:[{type:"text",text:li(a,o,`file:${t}:error`)}],isError:!0}:{content:[{type:"text",text:a}],isError:!0}}let i=s.stdout||"(no output)";return o&&o.trim().length>0&&Buffer.byteLength(i)>ui?{content:[{type:"text",text:li(i,o,`file:${t}`)}]}:{content:[{type:"text",text:i}]}}catch(s){return{content:[{type:"text",text:`Runtime error: ${s instanceof Error?s.message:String(s)}`}],isError:!0}}});dn.registerTool("index",{title:"Index Content",description:`Index documentation or knowledge content into a searchable BM25 knowledge base. Chunks markdown by headings (keeping code blocks intact) and stores in ephemeral FTS5 database. The full content does NOT stay in context \u2014 only a brief summary is returned.
 
 WHEN TO USE:
@@ -253,7 +257,7 @@ function decodeEntities(s) {
 main();
 `;dn.registerTool("fetch_and_index",{title:"Fetch & Index URL",description:`Fetches URL content, converts HTML to markdown, and indexes into the searchable knowledge base. Raw content never enters context \u2014 only a brief confirmation is returned.
 
-Use INSTEAD of WebFetch/Context7 when you need to reference web documentation later via search.
+PREFER THIS OVER WebFetch when you need to reference web documentation later via search. WebFetch loads entire page content into context; this tool indexes it and lets you search() on-demand.
 
 After fetching, use 'search' to retrieve specific sections on-demand.`,inputSchema:ie.object({url:ie.string().describe("The URL to fetch and index"),source:ie.string().optional().describe("Label for the indexed content (e.g., 'React useEffect docs', 'Supabase Auth API')")})},async({url:t,source:e})=>{try{let r=`process.argv[1] = ${JSON.stringify(t)};
 ${yT}`,n=await Xl.execute({language:"javascript",code:r,timeout:3e4});return n.exitCode!==0?{content:[{type:"text",text:`Failed to fetch ${t}: ${n.stderr||n.stdout}`}],isError:!0}:!n.stdout||n.stdout.trim().length===0?{content:[{type:"text",text:`Fetched ${t} but got empty content after HTML conversion`}],isError:!0}:_T(n.stdout,e??t)}catch(r){return{content:[{type:"text",text:`Fetch error: ${r instanceof Error?r.message:String(r)}`}],isError:!0}}});async function vT(){let t=new ii;await dn.connect(t),console.error(`Context Mode MCP server v${Qg} running on stdio`),console.error(`Detected runtimes: