diff --git a/.github/workflows/issue-monster.lock.yml b/.github/workflows/issue-monster.lock.yml
index 52a01c4bc2a..bfe28e24d60 100644
--- a/.github/workflows/issue-monster.lock.yml
+++ b/.github/workflows/issue-monster.lock.yml
@@ -1,4 +1,4 @@
-# gh-aw-metadata: {"schema_version":"v3","frontmatter_hash":"eac7305dc9f610adb3d3359f9367aca98730d48712f510c74e8523a3c70facf3","strict":true,"agent_id":"copilot","agent_model":"gpt-5.1-codex-mini"}
+# gh-aw-metadata: {"schema_version":"v3","frontmatter_hash":"d10076fe5d53fb3703246a470b8702dfe80f8d1d2f7f209456cb47e04446c174","strict":true,"agent_id":"copilot","agent_model":"claude-haiku-4.5"}
 # gh-aw-manifest: {"version":1,"secrets":["COPILOT_GITHUB_TOKEN","GH_AW_AGENT_TOKEN","GH_AW_GITHUB_MCP_SERVER_TOKEN","GH_AW_GITHUB_TOKEN","GITHUB_TOKEN"],"actions":[{"repo":"actions/checkout","sha":"de0fac2e4500dabe0009e67214ff5f5447ce83dd","version":"v6.0.2"},{"repo":"actions/download-artifact","sha":"3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c","version":"v8.0.1"},{"repo":"actions/github-script","sha":"373c709c69115d41ff229c7e5df9f8788daa9553","version":"v9"},{"repo":"actions/upload-artifact","sha":"043fb46d1a93c77aae656e7c1c64a875d1fc6a0a","version":"v7.0.1"}],"containers":[{"image":"ghcr.io/github/gh-aw-firewall/agent:0.25.24"},{"image":"ghcr.io/github/gh-aw-firewall/api-proxy:0.25.24"},{"image":"ghcr.io/github/gh-aw-firewall/squid:0.25.24"},{"image":"ghcr.io/github/gh-aw-mcpg:v0.2.24"},{"image":"ghcr.io/github/github-mcp-server:v1.0.0"},{"image":"node:lts-alpine","digest":"sha256:01743339035a5c3c11a373cd7c83aeab6ed1457b55da6a69e014a95ac4e4700b","pinned_image":"node:lts-alpine@sha256:01743339035a5c3c11a373cd7c83aeab6ed1457b55da6a69e014a95ac4e4700b"}]}
 #    ___                   _   _      
 #   / _ \                 | | (_)     
@@ -76,6 +76,8 @@ name: "Issue Monster"
     # with:
       # script: |
         # const { owner, repo } = context.repo;
+        # const MAX_ISSUES_WITH_BODY_CONTEXT = 8;
+        # const BODY_SNIPPET_MAX_LENGTH = 600;
         # 
         # try {
           # // Check for recent rate-limited PRs to avoid scheduling more work during rate limiting
@@ -145,6 +147,7 @@ name: "Issue Monster"
             # core.setOutput('issue_count', 0);
             # core.setOutput('issue_numbers', '');
             # core.setOutput('issue_list', '');
+            # core.setOutput('issue_context', '');
             # core.setOutput('has_issues', 'false');
             # return;
           # }
@@ -381,6 +384,7 @@ name: "Issue Monster"
                 # number: issue.number,
                 # title: issue.title,
                 # labels: issue.labels.map(l => l.name),
+                # body: issue.body,
                 # created_at: issue.created_at,
                 # score
               # };
@@ -392,6 +396,15 @@ name: "Issue Monster"
             # const labelStr = i.labels.length > 0 ? ` [${i.labels.join(', ')}]` : '';
             # return `#${i.number}: ${i.title}${labelStr} (score: ${i.score.toFixed(1)})`;
           # }).join('\n');
+        # 
+          # // Pre-fetch compact body context for top candidates so the agent can
+          # // triage without extra reads in most runs.
+          # const issueContext = scoredIssues.slice(0, MAX_ISSUES_WITH_BODY_CONTEXT).map(i => {
+            # const body = (i.body || '').replace(/\s+/g, ' ').trim();
+            # const bodySnippet = body.length > BODY_SNIPPET_MAX_LENGTH ? `${body.slice(0, BODY_SNIPPET_MAX_LENGTH)}…` : body;
+            # const labelStr = i.labels.length > 0 ? i.labels.join(', ') : 'none';
+            # return `#${i.number} | score=${i.score.toFixed(1)} | labels=${labelStr}\nTitle: ${i.title}\nBody: ${bodySnippet || '(no body)'}`;
+          # }).join('\n\n---\n\n');
           # 
           # const issueNumbers = scoredIssues.map(i => i.number).join(',');
           # 
@@ -403,6 +416,7 @@ name: "Issue Monster"
           # core.setOutput('issue_count', scoredIssues.length);
           # core.setOutput('issue_numbers', issueNumbers);
           # core.setOutput('issue_list', issueList);
+          # core.setOutput('issue_context', issueContext);
           # 
           # if (scoredIssues.length === 0) {
             # core.info('🍽️ No suitable candidate issues - the plate is empty!');
@@ -415,6 +429,7 @@ name: "Issue Monster"
           # core.setOutput('issue_count', 0);
           # core.setOutput('issue_numbers', '');
           # core.setOutput('issue_list', '');
+          # core.setOutput('issue_context', '');
           # core.setOutput('has_issues', 'false');
         # }
   workflow_dispatch:
@@ -468,7 +483,7 @@ jobs:
         env:
           GH_AW_INFO_ENGINE_ID: "copilot"
           GH_AW_INFO_ENGINE_NAME: "GitHub Copilot CLI"
-          GH_AW_INFO_MODEL: "gpt-5.1-codex-mini"
+          GH_AW_INFO_MODEL: "claude-haiku-4.5"
           GH_AW_INFO_VERSION: "1.0.21"
           GH_AW_INFO_AGENT_VERSION: "1.0.21"
           GH_AW_INFO_WORKFLOW_NAME: "Issue Monster"
@@ -537,6 +552,7 @@ jobs:
           GH_AW_GITHUB_REPOSITORY: ${{ github.repository }}
           GH_AW_GITHUB_RUN_ID: ${{ github.run_id }}
           GH_AW_GITHUB_WORKSPACE: ${{ github.workspace }}
+          GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_CONTEXT: ${{ needs.pre_activation.outputs.issue_context }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_COUNT: ${{ needs.pre_activation.outputs.issue_count }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_LIST: ${{ needs.pre_activation.outputs.issue_list }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_NUMBERS: ${{ needs.pre_activation.outputs.issue_numbers }}
@@ -544,20 +560,20 @@ jobs:
         run: |
           bash "${RUNNER_TEMP}/gh-aw/actions/create_prompt_first.sh"
           {
-          cat << 'GH_AW_PROMPT_f8901b6202eb9fe6_EOF'
+          cat << 'GH_AW_PROMPT_e301ff5ec303eb90_EOF'
           <system>
-          GH_AW_PROMPT_f8901b6202eb9fe6_EOF
+          GH_AW_PROMPT_e301ff5ec303eb90_EOF
           cat "${RUNNER_TEMP}/gh-aw/prompts/xpia.md"
           cat "${RUNNER_TEMP}/gh-aw/prompts/temp_folder_prompt.md"
           cat "${RUNNER_TEMP}/gh-aw/prompts/markdown.md"
           cat "${RUNNER_TEMP}/gh-aw/prompts/safe_outputs_prompt.md"
-          cat << 'GH_AW_PROMPT_f8901b6202eb9fe6_EOF'
+          cat << 'GH_AW_PROMPT_e301ff5ec303eb90_EOF'
           <safe-output-tools>
           Tools: add_comment(max:3), assign_to_agent(max:3), missing_tool, missing_data, noop
           </safe-output-tools>
-          GH_AW_PROMPT_f8901b6202eb9fe6_EOF
+          GH_AW_PROMPT_e301ff5ec303eb90_EOF
           cat "${RUNNER_TEMP}/gh-aw/prompts/mcp_cli_tools_prompt.md"
-          cat << 'GH_AW_PROMPT_f8901b6202eb9fe6_EOF'
+          cat << 'GH_AW_PROMPT_e301ff5ec303eb90_EOF'
           <github-context>
           The following GitHub context information is available for this workflow:
           {{#if __GH_AW_GITHUB_ACTOR__ }}
@@ -586,20 +602,21 @@ jobs:
           {{/if}}
           </github-context>
           
-          GH_AW_PROMPT_f8901b6202eb9fe6_EOF
+          GH_AW_PROMPT_e301ff5ec303eb90_EOF
           cat "${RUNNER_TEMP}/gh-aw/prompts/github_mcp_tools_with_safeoutputs_prompt.md"
-          cat << 'GH_AW_PROMPT_f8901b6202eb9fe6_EOF'
+          cat << 'GH_AW_PROMPT_e301ff5ec303eb90_EOF'
           </system>
           {{#runtime-import .github/workflows/shared/github-guard-policy.md}}
           {{#runtime-import .github/workflows/shared/activation-app.md}}
           {{#runtime-import .github/workflows/issue-monster.md}}
-          GH_AW_PROMPT_f8901b6202eb9fe6_EOF
+          GH_AW_PROMPT_e301ff5ec303eb90_EOF
           } > "$GH_AW_PROMPT"
       - name: Interpolate variables and render templates
         uses: actions/github-script@373c709c69115d41ff229c7e5df9f8788daa9553 # v9
         env:
           GH_AW_PROMPT: /tmp/gh-aw/aw-prompts/prompt.txt
           GH_AW_GITHUB_REPOSITORY: ${{ github.repository }}
+          GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_CONTEXT: ${{ needs.pre_activation.outputs.issue_context }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_COUNT: ${{ needs.pre_activation.outputs.issue_count }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_LIST: ${{ needs.pre_activation.outputs.issue_list }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_NUMBERS: ${{ needs.pre_activation.outputs.issue_numbers }}
@@ -623,6 +640,7 @@ jobs:
           GH_AW_GITHUB_WORKSPACE: ${{ github.workspace }}
           GH_AW_MCP_CLI_SERVERS_LIST: '- `safeoutputs` — run `safeoutputs --help` to see available tools'
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ACTIVATED: ${{ needs.pre_activation.outputs.activated }}
+          GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_CONTEXT: ${{ needs.pre_activation.outputs.issue_context }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_COUNT: ${{ needs.pre_activation.outputs.issue_count }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_LIST: ${{ needs.pre_activation.outputs.issue_list }}
           GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_NUMBERS: ${{ needs.pre_activation.outputs.issue_numbers }}
@@ -647,6 +665,7 @@ jobs:
                 GH_AW_GITHUB_WORKSPACE: process.env.GH_AW_GITHUB_WORKSPACE,
                 GH_AW_MCP_CLI_SERVERS_LIST: process.env.GH_AW_MCP_CLI_SERVERS_LIST,
                 GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ACTIVATED: process.env.GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ACTIVATED,
+                GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_CONTEXT: process.env.GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_CONTEXT,
                 GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_COUNT: process.env.GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_COUNT,
                 GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_LIST: process.env.GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_LIST,
                 GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_NUMBERS: process.env.GH_AW_NEEDS_PRE_ACTIVATION_OUTPUTS_ISSUE_NUMBERS
@@ -782,9 +801,9 @@ jobs:
           mkdir -p "${RUNNER_TEMP}/gh-aw/safeoutputs"
           mkdir -p /tmp/gh-aw/safeoutputs
           mkdir -p /tmp/gh-aw/mcp-logs/safeoutputs
-          cat > "${RUNNER_TEMP}/gh-aw/safeoutputs/config.json" << 'GH_AW_SAFE_OUTPUTS_CONFIG_c922b5e93a111615_EOF'
+          cat > "${RUNNER_TEMP}/gh-aw/safeoutputs/config.json" << 'GH_AW_SAFE_OUTPUTS_CONFIG_8c9d5b9222cb4f50_EOF'
           {"add_comment":{"max":3,"target":"*"},"assign_to_agent":{"allowed":["copilot"],"max":3,"target":"*"},"create_report_incomplete_issue":{},"missing_data":{},"missing_tool":{},"noop":{"max":1,"report-as-issue":"true"},"report_incomplete":{}}
-          GH_AW_SAFE_OUTPUTS_CONFIG_c922b5e93a111615_EOF
+          GH_AW_SAFE_OUTPUTS_CONFIG_8c9d5b9222cb4f50_EOF
       - name: Write Safe Outputs Tools
         env:
           GH_AW_TOOLS_META_JSON: |
@@ -996,7 +1015,7 @@ jobs:
           
           mkdir -p /home/runner/.copilot
           GH_AW_NODE=$(which node 2>/dev/null || command -v node 2>/dev/null || echo node)
-          cat << GH_AW_MCP_CONFIG_f578c422c31b5a81_EOF | "$GH_AW_NODE" "${RUNNER_TEMP}/gh-aw/actions/start_mcp_gateway.cjs"
+          cat << GH_AW_MCP_CONFIG_214645539377f980_EOF | "$GH_AW_NODE" "${RUNNER_TEMP}/gh-aw/actions/start_mcp_gateway.cjs"
           {
             "mcpServers": {
               "github": {
@@ -1006,7 +1025,7 @@ jobs:
                   "GITHUB_HOST": "\${GITHUB_SERVER_URL}",
                   "GITHUB_PERSONAL_ACCESS_TOKEN": "\${GITHUB_MCP_SERVER_TOKEN}",
                   "GITHUB_READ_ONLY": "1",
-                  "GITHUB_TOOLSETS": "context,repos,issues,pull_requests"
+                  "GITHUB_TOOLSETS": "issues"
                 },
                 "guard-policies": {
                   "allow-only": {
@@ -1040,7 +1059,7 @@ jobs:
               "payloadDir": "${MCP_GATEWAY_PAYLOAD_DIR}"
             }
           }
-          GH_AW_MCP_CONFIG_f578c422c31b5a81_EOF
+          GH_AW_MCP_CONFIG_214645539377f980_EOF
       - name: Mount MCP servers as CLIs
         id: mount-mcp-clis
         continue-on-error: true
@@ -1085,7 +1104,7 @@ jobs:
         env:
           COPILOT_AGENT_RUNNER_TYPE: STANDALONE
           COPILOT_GITHUB_TOKEN: ${{ secrets.COPILOT_GITHUB_TOKEN }}
-          COPILOT_MODEL: gpt-5.1-codex-mini
+          COPILOT_MODEL: claude-haiku-4.5
           GH_AW_MCP_CONFIG: /home/runner/.copilot/mcp-config.json
           GH_AW_PHASE: agent
           GH_AW_PROMPT: /tmp/gh-aw/aw-prompts/prompt.txt
@@ -1529,7 +1548,7 @@ jobs:
         env:
           COPILOT_AGENT_RUNNER_TYPE: STANDALONE
           COPILOT_GITHUB_TOKEN: ${{ secrets.COPILOT_GITHUB_TOKEN }}
-          COPILOT_MODEL: gpt-5.1-codex-mini
+          COPILOT_MODEL: claude-haiku-4.5
           GH_AW_PHASE: detection
           GH_AW_PROMPT: /tmp/gh-aw/aw-prompts/prompt.txt
           GH_AW_VERSION: dev
@@ -1576,6 +1595,7 @@ jobs:
     outputs:
       activated: ${{ steps.check_membership.outputs.is_team_member == 'true' && steps.check_skip_if_match.outputs.skip_check_ok == 'true' && steps.check_skip_if_no_match.outputs.skip_no_match_check_ok == 'true' && steps.check_skip_if_check_failing.outputs.skip_if_check_failing_ok == 'true' }}
       has_issues: ${{ steps.search.outputs.has_issues }}
+      issue_context: ${{ steps.search.outputs.issue_context }}
       issue_count: ${{ steps.search.outputs.issue_count }}
       issue_list: ${{ steps.search.outputs.issue_list }}
       issue_numbers: ${{ steps.search.outputs.issue_numbers }}
@@ -1652,6 +1672,8 @@ jobs:
         with:
           script: |
             const { owner, repo } = context.repo;
+            const MAX_ISSUES_WITH_BODY_CONTEXT = 8;
+            const BODY_SNIPPET_MAX_LENGTH = 600;
 
             try {
               // Check for recent rate-limited PRs to avoid scheduling more work during rate limiting
@@ -1721,6 +1743,7 @@ jobs:
                 core.setOutput('issue_count', 0);
                 core.setOutput('issue_numbers', '');
                 core.setOutput('issue_list', '');
+                core.setOutput('issue_context', '');
                 core.setOutput('has_issues', 'false');
                 return;
               }
@@ -1957,6 +1980,7 @@ jobs:
                     number: issue.number,
                     title: issue.title,
                     labels: issue.labels.map(l => l.name),
+                    body: issue.body,
                     created_at: issue.created_at,
                     score
                   };
@@ -1969,6 +1993,15 @@ jobs:
                 return `#${i.number}: ${i.title}${labelStr} (score: ${i.score.toFixed(1)})`;
               }).join('\n');
 
+              // Pre-fetch compact body context for top candidates so the agent can
+              // triage without extra reads in most runs.
+              const issueContext = scoredIssues.slice(0, MAX_ISSUES_WITH_BODY_CONTEXT).map(i => {
+                const body = (i.body || '').replace(/\s+/g, ' ').trim();
+                const bodySnippet = body.length > BODY_SNIPPET_MAX_LENGTH ? `${body.slice(0, BODY_SNIPPET_MAX_LENGTH)}…` : body;
+                const labelStr = i.labels.length > 0 ? i.labels.join(', ') : 'none';
+                return `#${i.number} | score=${i.score.toFixed(1)} | labels=${labelStr}\nTitle: ${i.title}\nBody: ${bodySnippet || '(no body)'}`;
+              }).join('\n\n---\n\n');
+
               const issueNumbers = scoredIssues.map(i => i.number).join(',');
 
               core.info(`Total candidate issues after filtering: ${scoredIssues.length}`);
@@ -1979,6 +2012,7 @@ jobs:
               core.setOutput('issue_count', scoredIssues.length);
               core.setOutput('issue_numbers', issueNumbers);
               core.setOutput('issue_list', issueList);
+              core.setOutput('issue_context', issueContext);
 
               if (scoredIssues.length === 0) {
                 core.info('🍽️ No suitable candidate issues - the plate is empty!');
@@ -1991,6 +2025,7 @@ jobs:
               core.setOutput('issue_count', 0);
               core.setOutput('issue_numbers', '');
               core.setOutput('issue_list', '');
+              core.setOutput('issue_context', '');
               core.setOutput('has_issues', 'false');
             }
 
@@ -2013,7 +2048,7 @@ jobs:
       GH_AW_DETECTION_REASON: ${{ needs.detection.outputs.detection_reason }}
       GH_AW_EFFECTIVE_TOKENS: ${{ needs.agent.outputs.effective_tokens }}
       GH_AW_ENGINE_ID: "copilot"
-      GH_AW_ENGINE_MODEL: "gpt-5.1-codex-mini"
+      GH_AW_ENGINE_MODEL: "claude-haiku-4.5"
       GH_AW_SAFE_OUTPUT_MESSAGES: "{\"footer\":\"\\u003e 🍪 *Om nom nom by [{workflow_name}]({run_url})*{effective_tokens_suffix}{history_link}\",\"runStarted\":\"🍪 ISSUE! ISSUE! [{workflow_name}]({run_url}) hungry for issues on this {event_type}! Om nom nom...\",\"runSuccess\":\"🍪 YUMMY! [{workflow_name}]({run_url}) ate the issues! That was DELICIOUS! Me want MORE! 😋\",\"runFailure\":\"🍪 Aww... [{workflow_name}]({run_url}) {status}. No cookie for monster today... 😢\"}"
       GH_AW_WORKFLOW_ID: "issue-monster"
       GH_AW_WORKFLOW_NAME: "Issue Monster"
diff --git a/.github/workflows/issue-monster.md b/.github/workflows/issue-monster.md
index bd896a1f050..88f63cc4cfc 100644
--- a/.github/workflows/issue-monster.md
+++ b/.github/workflows/issue-monster.md
@@ -25,6 +25,8 @@ on:
       with:
         script: |
           const { owner, repo } = context.repo;
+          const MAX_ISSUES_WITH_BODY_CONTEXT = 8;
+          const BODY_SNIPPET_MAX_LENGTH = 600;
           
           try {
             // Check for recent rate-limited PRs to avoid scheduling more work during rate limiting
@@ -94,6 +96,7 @@ on:
               core.setOutput('issue_count', 0);
               core.setOutput('issue_numbers', '');
               core.setOutput('issue_list', '');
+              core.setOutput('issue_context', '');
               core.setOutput('has_issues', 'false');
               return;
             }
@@ -330,6 +333,7 @@ on:
                   number: issue.number,
                   title: issue.title,
                   labels: issue.labels.map(l => l.name),
+                  body: issue.body,
                   created_at: issue.created_at,
                   score
                 };
@@ -341,6 +345,15 @@ on:
               const labelStr = i.labels.length > 0 ? ` [${i.labels.join(', ')}]` : '';
               return `#${i.number}: ${i.title}${labelStr} (score: ${i.score.toFixed(1)})`;
             }).join('\n');
+
+            // Pre-fetch compact body context for top candidates so the agent can
+            // triage without extra reads in most runs.
+            const issueContext = scoredIssues.slice(0, MAX_ISSUES_WITH_BODY_CONTEXT).map(i => {
+              const body = (i.body || '').replace(/\s+/g, ' ').trim();
+              const bodySnippet = body.length > BODY_SNIPPET_MAX_LENGTH ? `${body.slice(0, BODY_SNIPPET_MAX_LENGTH)}…` : body;
+              const labelStr = i.labels.length > 0 ? i.labels.join(', ') : 'none';
+              return `#${i.number} | score=${i.score.toFixed(1)} | labels=${labelStr}\nTitle: ${i.title}\nBody: ${bodySnippet || '(no body)'}`;
+            }).join('\n\n---\n\n');
             
             const issueNumbers = scoredIssues.map(i => i.number).join(',');
             
@@ -352,6 +365,7 @@ on:
             core.setOutput('issue_count', scoredIssues.length);
             core.setOutput('issue_numbers', issueNumbers);
             core.setOutput('issue_list', issueList);
+            core.setOutput('issue_context', issueContext);
             
             if (scoredIssues.length === 0) {
               core.info('🍽️ No suitable candidate issues - the plate is empty!');
@@ -364,6 +378,7 @@ on:
             core.setOutput('issue_count', 0);
             core.setOutput('issue_numbers', '');
             core.setOutput('issue_list', '');
+            core.setOutput('issue_context', '');
             core.setOutput('has_issues', 'false');
           }
 
@@ -375,7 +390,7 @@ permissions:
 
 engine:
   id: copilot
-  model: gpt-5.1-codex-mini
+  model: claude-haiku-4.5
 
 imports:
   - shared/github-guard-policy.md
@@ -387,7 +402,7 @@ tools:
   mount-as-clis: true
   github:
     min-integrity: approved
-    toolsets: [default, pull_requests]
+    toolsets: [issues]
 
 if: needs.pre_activation.outputs.has_issues == 'true'
 
@@ -397,6 +412,7 @@ jobs:
       issue_count: ${{ steps.search.outputs.issue_count }}
       issue_numbers: ${{ steps.search.outputs.issue_numbers }}
       issue_list: ${{ steps.search.outputs.issue_list }}
+      issue_context: ${{ steps.search.outputs.issue_context }}
       has_issues: ${{ steps.search.outputs.has_issues }}
 
 safe-outputs:
@@ -473,7 +489,12 @@ Issues are scored and sorted by priority:
 ${{ needs.pre_activation.outputs.issue_list }}
 ```
 
-Work with this pre-fetched, filtered, and prioritized list of issues. Do not perform additional searches - the issue numbers are already identified above, sorted from highest to lowest priority.
+**Pre-fetched Body Context (top candidates):**
+```
+${{ needs.pre_activation.outputs.issue_context }}
+```
+
+Work with this pre-fetched, filtered, and prioritized list of issues. Do not perform additional searches - candidate issue numbers and body excerpts are already identified above.
 
 ### 1a. Handle Parent-Child Issue Relationships (for "task" or "plan" labeled issues)
 
@@ -494,17 +515,7 @@ For issues with the "task" or "plan" label, check if they are sub-issues linked
 - Only after #101's PR is merged/closed, process #102
 - This ensures orderly, sequential processing of related tasks
 
-### 2. Review the Pre-Filtered Issue List
-
-The pre-activation job has already performed comprehensive filtering, including:
-- Issues already assigned to Copilot
-- Issues with open PRs linked to them (from any author)
-- Issues with closed/merged PRs (treated as complete)
-- **For "task" or "plan" labeled sub-issues**: Check if any sibling sub-issue (same parent) has an open PR from Copilot
-
-The list you receive has already been filtered to exclude all of these cases, so you can focus on the actual assignment logic.
-
-### 3. Select Up to Three Issues to Work On
+### 2. Select Up to Three Issues to Work On
 
 From the prioritized and filtered list (issues WITHOUT Copilot assignments or open PRs):
 - **Select up to three appropriate issues** to assign
@@ -538,10 +549,13 @@ From the prioritized and filtered list (issues WITHOUT Copilot assignments or op
 - Assign only the issues that are clearly separate in topic
 - Do not force assignments just to reach the maximum
 
-### 4. Read and Understand Each Selected Issue
+### 3. Validate Selected Issues (Body-First)
 
 For each selected issue (which has already been pre-filtered to ensure no open/closed PRs exist):
-- Read the full issue body and any comments
+- Use the pre-fetched body context first
+- If a body excerpt is ambiguous, call `issue_read` with `method: get` for that issue
+- Do **not** fetch comments by default
+- Only fetch comments (`issue_read` with `method: get_comments`) when a specific triage rule truly requires comment context (for example: to confirm whether maintainers already requested a specific implementation approach, or to capture additional repro steps posted after the original issue body)
 - Understand what fix is needed
 - Identify the files that need to be modified
 - Verify it doesn't overlap with the other selected issues
@@ -562,7 +576,7 @@ Some issues may be blocked by an integrity policy when you try to read them with
 → Call `noop` with: `"🛡️ All 3 candidates (#100, #102, #105) were integrity-filtered. No assignments made this run."`
 
 
-### 5. Assign Issues to Copilot Agent
+### 4. Assign Issues to Copilot Agent
 
 For each selected issue, use the `assign_to_agent` tool from the `safeoutputs` MCP server to assign the Copilot coding agent:
 
@@ -578,7 +592,7 @@ The Copilot coding agent will:
 3. Create a pull request with the fix
 4. Follow the repository's AGENTS.md guidelines
 
-### 6. Add Comment to Each Assigned Issue
+### 5. Add Comment to Each Assigned Issue
 
 For each issue you assign, use the `add_comment` tool from the `safeoutputs` MCP server to add a comment:
 
@@ -616,21 +630,11 @@ Issue Monster runs frequently (every 30 minutes), so keeping each run lean is cr
 ## Success Criteria
 
 A successful run means:
-1. **Rate limiting check passed** - The search verified no recent PRs are rate-limited (or workflow skipped if rate limiting detected)
-2. You reviewed the pre-searched, filtered, and prioritized issue list
-3. The search already excluded issues with problematic labels (wontfix, question, discussion, etc.)
-4. The search already excluded issues with campaign labels (campaign:*) as these are managed by campaign orchestrators
-5. The search already excluded issues that already have assignees
-6. The search already excluded issues that have sub-issues (parent/organizing issues are not tasks)
-7. The search already excluded issues with closed or merged PRs (treated as complete)
-8. The search already excluded issues with open PRs from Copilot coding agent (already being worked on)
-9. Issues are sorted by priority score (good-first-issue, bug, security, etc. get higher scores)
-10. For "task" or "plan" issues: You checked for parent issues and sibling sub-issue PRs if necessary
-11. You selected up to three appropriate issues from the top of the priority list that are completely separate in topic
-12. You read and understood each issue
-13. You verified that the selected issues don't have overlapping concerns or file changes
-14. You assigned each issue to the Copilot coding agent using `assign_to_agent`
-15. You commented on each issue being assigned
+1. You used the pre-fetched prioritized list (and body context) without re-searching
+2. You selected up to three issues that are clearly separate in topic
+3. You used body-first validation and only fetched comments when strictly necessary
+4. You assigned each selected issue to Copilot using `assign_to_agent`
+5. You commented on each assigned issue (or called `noop` when no assignments were made)
 
 ## Error Handling