diff --git a/docs/blog/2026-04-23-platform-instructions-governance/index.md b/docs/blog/2026-04-23-platform-instructions-governance/index.md
index 959d80da..872d8ae3 100644
--- a/docs/blog/2026-04-23-platform-instructions-governance/index.md
+++ b/docs/blog/2026-04-23-platform-instructions-governance/index.md
@@ -8,6 +8,7 @@ og_description: "Platform Instructions: global and workspace-scoped rules prepen
 tags: [governance, platform-instructions, enterprise, security, it-governance, system-prompt, policy, a2a]
 keywords: [AI fleet governance, enterprise AI policy, system prompt governance, AI agent compliance, platform instructions, workspace policy enforcement, enterprise AI security, AI agent ACL]
 canonical: https://docs.molecule.ai/blog/govern-ai-fleet-system-prompt-level
+og_image: ""
 ---
 
 <script type="application/ld+json">
@@ -99,6 +100,7 @@ Platform Instructions are available on **Enterprise plans**. To get started:
 - Verify resolved instructions via `GET /workspaces/{id}/instructions/resolve`
 
 For a complete governance picture, combine Platform Instructions with [Tool Trace](/blog/ai-agent-observability-without-overhead/) — see exactly which tools were called and what inputs were passed, alongside the policy that governed them.
+- See both features in one post: [Tool Trace + Platform Instructions Overview](/blog/tool-trace-platform-instructions-overview/)
 
 ---
 
diff --git a/docs/blog/2026-04-23-tool-trace-observability/index.md b/docs/blog/2026-04-23-tool-trace-observability/index.md
index e89f45c2..9327b49d 100644
--- a/docs/blog/2026-04-23-tool-trace-observability/index.md
+++ b/docs/blog/2026-04-23-tool-trace-observability/index.md
@@ -8,6 +8,7 @@ og_description: "See every tool your agent called — inputs, outputs, timing 
 tags: [observability, tool-trace, debugging, devops, platform-engineering, a2a, claude]
 keywords: [AI agent observability, tool trace debugging, Claude agent debugging, agent audit trail, parallel tool call trace, run_id pairing, AI agent monitoring, DevOps agent observability]
 canonical: https://docs.molecule.ai/blog/ai-agent-observability-without-overhead
+og_image: ""
 ---
 
 <script type="application/ld+json">
@@ -106,6 +107,7 @@ Combined with the [org-scoped API key audit trail](/docs/blog/2026-04-21-org-sco
 - Query `activity_logs.tool_trace` JSONB for historical traces
 - Combine with org API key attribution for complete fleet observability
 - Read the [A2A protocol documentation](/docs/api-protocol/a2a-protocol.md)
+- Govern agent behavior with [Platform Instructions — system-prompt level policy](/blog/govern-ai-fleet-system-prompt-level/)
 
 ---
 
diff --git a/docs/blog/2026-04-23-tool-trace-platform-instructions/index.md b/docs/blog/2026-04-23-tool-trace-platform-instructions/index.md
index 160cd33c..4068eb7b 100644
--- a/docs/blog/2026-04-23-tool-trace-platform-instructions/index.md
+++ b/docs/blog/2026-04-23-tool-trace-platform-instructions/index.md
@@ -1,13 +1,14 @@
 ---
 title: "Tool Trace + Platform Instructions: Full Visibility and Policy-Level Governance"
 date: 2026-04-23
-slug: tool-trace-platform-instructions
+slug: tool-trace-platform-instructions-overview
 description: "See every tool your agent called — inputs, outputs, timing — in real-time. And enforce org-wide governance policy at the system prompt level with Platform Instructions."
 og_title: "Tool Trace + Platform Instructions: Full Visibility and Policy-Level Governance"
 og_description: "Tool-level observability in every A2A response meets system-prompt governance. Two enterprise-grade features, shipped together."
 tags: [tool-trace, observability, platform-instructions, governance, enterprise, debugging, a2a]
 keywords: [AI agent debugging, tool trace observability, agent governance, platform instructions, enterprise AI audit, system prompt governance, Claude tool call visibility, agent observability]
-canonical: https://docs.molecule.ai/blog/tool-trace-platform-instructions
+canonical: https://docs.molecule.ai/blog/tool-trace-platform-instructions-overview
+og_image: ""
 ---
 
 <script type="application/ld+json">
@@ -116,7 +117,9 @@ When something goes wrong — an agent calls an unexpected tool, or behavior dri
 
 - Tool Trace is enabled by default on all workspaces. Check `Message.metadata.tool_trace` in your A2A responses.
 - Platform Instructions are available on Enterprise plans. Visit your workspace settings or contact your account team.
-- Explore the full A2A protocol documentation in `docs/api-protocol/a2a-protocol.md`.
+- Explore the [A2A protocol documentation](/docs/api-protocol/a2a-protocol.md)
+- Deep-dive: [Tool Trace — every tool call captured](/blog/ai-agent-observability-without-overhead/)
+- Deep-dive: [Platform Instructions — governance at the system prompt level](/blog/govern-ai-fleet-system-prompt-level/)
 
 ---
 
diff --git a/marketing/devrel/phase34-platform-instructions-social-copy.md b/marketing/devrel/phase34-platform-instructions-social-copy.md
new file mode 100644
index 00000000..dc94639c
--- /dev/null
+++ b/marketing/devrel/phase34-platform-instructions-social-copy.md
@@ -0,0 +1,19 @@
+---
+title: "Platform Instructions: Governance at the System Prompt Level"
+date: 2026-04-23
+platform: linkedin
+author: Molecule AI
+og_image: ""
+---
+
+Enterprise governance before the first agent turn executes.
+
+Platform Instructions: write policy rules once, enforce them everywhere — at the system prompt level.
+
+Global + workspace-scoped rules. wsAuth-gated resolve. No code deploys.
+
+Available on Enterprise plans.
+
+→ docs.molecule.ai/blog/govern-ai-fleet-system-prompt-level
+
+#AI #Enterprise #Governance #Security #Agents
\ No newline at end of file
diff --git a/marketing/devrel/phase34-tool-trace-social-copy.md b/marketing/devrel/phase34-tool-trace-social-copy.md
new file mode 100644
index 00000000..f2520480
--- /dev/null
+++ b/marketing/devrel/phase34-tool-trace-social-copy.md
@@ -0,0 +1,19 @@
+---
+title: "Tool Trace: Debugging Agents Like a Pro"
+date: 2026-04-23
+platform: x
+author: Molecule AI
+og_image: ""
+---
+
+🔍 Tool Trace — now in every A2A response.
+
+See every tool your agent called → inputs, output previews, timing. Parallel calls handled correctly via run_id pairing.
+
+No sampling. No sidecar. No guesswork.
+
+Built into the protocol. Enabled by default.
+
+→ docs.molecule.ai/blog/ai-agent-observability-without-overhead
+
+#AI #DevOps #Agents #Observability #Claude
\ No newline at end of file