diff --git a/README.md b/README.md index 920753a..066425e 100644 --- a/README.md +++ b/README.md @@ -43,6 +43,7 @@ uvx --from git+https://github.com/liatrio-labs/slash-command-manager \ - **Prompt-first workflow:** Use curated prompts to go from idea → spec → task list → implementation-ready backlog. - **Predictable delivery:** Every step emphasizes demoable slices, proof artifacts, and collaboration with junior developers in mind. - **No dependencies required:** The prompts are plain Markdown files that work with any AI assistant. +- **Context verification:** Built-in emoji markers (SDD1️⃣-SDD4️⃣) detect when AI responses follow critical instructions, helping identify context rot issues early. ## Why Spec-Driven Development? @@ -67,6 +68,14 @@ All prompts live in `prompts/` and are designed for use inside your preferred AI Each prompt writes Markdown outputs into `docs/specs/[NN]-spec-[feature-name]/` (where `[NN]` is a zero-padded 2-digit number: 01, 02, 03, etc.), giving you a lightweight backlog that is easy to review, share, and implement. +### Context Verification Markers + +Each prompt includes a context verification marker (SDD1️⃣ for spec generation, SDD2️⃣ for task breakdown, SDD3️⃣ for task management, SDD4️⃣ for validation) that appears at the start of AI responses. These markers help detect **context rot**—a phenomenon where AI performance degrades as input context length increases, even when tasks remain simple. + +**Why this matters:** Context rot doesn't announce itself with errors. It creeps in silently, causing models to lose track of critical instructions. When you see the marker at the start of each response, it's an indicator that the AI is probably following the prompt's instructions. If the marker disappears, it's an immediate signal that context instructions may have been lost. + +**What to expect:** You'll see responses like `SDD1️⃣ I'll help you generate a specification...` or `SDD3️⃣ Let me start implementing task 1.0...`. This is normal and indicates the verification system is working. For more details, see the [research documentation](docs/emoji-context-verification-research.md). + ## How does it work? The workflow is driven by Markdown prompts that function as reusable playbooks for the AI agent. Reference the prompts directly, or install them as slash commands using the [slash-command-manager](https://github.com/liatrio-labs/slash-command-manager), to keep the AI focused on structured outcomes. diff --git a/docs/common-questions.html b/docs/common-questions.html index 613ca43..93b6f5e 100644 --- a/docs/common-questions.html +++ b/docs/common-questions.html @@ -302,6 +302,90 @@

The SDD Advantage

+ + +
+
+

Why Do AI Responses Start with Emoji Markers (SDD1️⃣, SDD2️⃣, etc.)?

+

You may notice that AI responses begin with emoji markers like + SDD1️⃣, SDD2️⃣, SDD3️⃣, or SDD4️⃣. This is an + intentional feature designed to detect a silent failure mode called context rot. +

+ +
+
+
+ +
+

What Is Context Rot?

+

Research from Chroma and Anthropic demonstrates that AI performance degrades as input context + length increases, even when tasks remain simple. This degradation happens silently—the AI + doesn't announce errors, but gradually loses track of critical instructions.

+
+ +
+
+ +
+

How Verification Markers Work

+

Each prompt instructs the AI to always begin responses with its specific marker (SDD1️⃣ for + spec generation, SDD2️⃣ for task breakdown, etc.). When you see the marker, it's an + indicator that critical instructions are probably being followed. If the + marker disappears, it's an immediate signal that context instructions may have been lost. +

+
+ +
+
+ +
+

What You Should Expect

+

Normal responses will start with the marker: + SDD1️⃣ I'll help you generate a specification... or + SDD3️⃣ Let me start implementing task 1.0.... This is expected behavior and + indicates the verification system is working correctly. The markers add minimal overhead + (1-2 tokens) while providing immediate visual feedback. +

+
+
+ +
+

Technical Background

+
+

This verification technique was shared by Lada Kesseler at AI Native Dev Con Fall 2025 as a + practical solution for detecting context rot in production AI workflows. The technique + provides:

+
    +
  • Immediate feedback: Visual confirmation that instructions are being + followed
    • +
  • Low overhead: Minimal token cost (1-2 tokens per response)
    • +
  • Simple implementation: Easy to spot in terminal/text output
    • +
  • Failure detection: Absence of marker immediately signals instruction + loss
    • +
+

For detailed research and technical information, see the context verification research + documentation.

+
+
+
+