WFGY/ProblemMap/GlobalFixMap/Agents_Orchestration/langchain.md

LangChain — Guardrails and Fix Patterns

🧭 Quick Return to Map

You are in a sub-page of Agents & Orchestration.
To reorient, go back here:

• Agents & Orchestration — orchestration frameworks and guardrails
• WFGY Global Fix Map — main Emergency Room, 300+ structured fixes
• WFGY Problem Map 1.0 — 16 reproducible failure modes

Think of this page as a desk within a ward.
If you need the full triage and all prescriptions, return to the Emergency Room lobby.

Use this when your pipeline is built with LangChain (LCEL, Runnable*, Agents, Tools) and you see wrong snippets, unstable reasoning, mixed sources, or silent failures that look fine in logs.

Acceptance targets

• ΔS(question, retrieved) ≤ 0.45
• coverage ≥ 0.70 to the intended section or record
• λ stays convergent across 3 paraphrases

---

Typical breakpoints → exact fixes

• Retrieval returns plausible but wrong chunks
  Fix No.1: Hallucination & Chunk Drift →
  Hallucination
  Also review Retrieval Playbook →
  Retrieval Playbook

• High vector similarity, wrong meaning
  Fix No.5: Embedding ≠ Semantic →
  Embedding ≠ Semantic

• Hybrid chains with BM25 or HyDE degrade compared to single retriever
  Pattern: Query Parsing Split →
  Query Parsing Split
  Apply Rerankers for ordering control →
  Rerankers

• Facts exist in the store but never show up
  Pattern: Vectorstore Fragmentation →
  Vectorstore Fragmentation

• Citations missing or inconsistent across LCEL steps
  Fix No.8: Retrieval Traceability + Data Contracts →
  Retrieval Traceability · Data Contracts

• Long Runnable chains flatten style and drift logically
  Fix No.3 and No.9: Context Drift and Entropy Collapse →
  Context Drift · Entropy Collapse

• Agent loops, role confusion, memory overwrite
  Fix No.13: Multi-Agent Chaos →
  Multi-Agent Problems · Role Drift · Memory Overwrite

• Model sounds confident but is wrong
  Fix No.4: Bluffing / Overconfidence →
  Bluffing

• Model merges two sources into one answer
  Pattern: Symbolic Constraint Unlock →
  Symbolic Constraint Unlock

---

Minimal LCEL pattern with WFGY checks

# Pseudocode. Show the control points you must keep.
from langchain_core.runnables import RunnablePassthrough, RunnableMap

def retrieve(q):
    # k sweep and unified analyzer across dense and sparse
    return retriever.invoke(q, k=10)

def assemble(context, q):
    # schema-locked: system -> task -> constraints -> citations -> answer
    return prompt.format(context=context, question=q)

def reason(msg):
    # model call runs after cite-then-explain requirement in the prompt
    return llm.invoke(msg)

def wfgy_checks(q, context, answer):
    # compute ΔS(question, context) and trace why this snippet
    # enforce thresholds and fail fast when ΔS ≥ 0.60 or λ divergent
    return metrics_and_trace(q, context, answer)

chain = (
    {"q": RunnablePassthrough()}
    | RunnableMap({"context": lambda x: retrieve(x["q"]), "q": lambda x: x["q"]})
    | RunnableMap({"msg": lambda x: assemble(x["context"], x["q"]), "q": lambda x: x["q"], "context": lambda x: x["context"]})
    | RunnableMap({"answer": lambda x: reason(x["msg"]), "q": lambda x: x["q"], "context": lambda x: x["context"]})
    | (lambda x: wfgy_checks(x["q"], x["context"], x["answer"]))
)

What this enforces

• Retrieval is observable and parameterized.
• Prompt is schema locked with cite first.
• WFGY check runs after generation and can stop the run when ΔS is high or λ flips.
• Traces record snippet to citation mapping for audits.

Specs and recipes RAG Architecture & Recovery · Retrieval Playbook · Retrieval Traceability · Data Contracts

---

LangChain-specific gotchas

• Mixed embedding functions across write and read paths. Rebuild with explicit metric and normalization. See Embedding ≠ Semantic

• RunnableParallel merges outputs without source fences. Add per-source headers and forbid cross-section reuse. See Symbolic Constraint Unlock

• Memory modules re-assert old facts after a refresh. Stamp mem_rev and mem_hash. See Memory Desync

• Agents tool-call retry loops. Add BBCR bridge steps and clamp variance with BBAM in the prompt recipe. See Logic Collapse

---

When to escalate

• ΔS remains ≥ 0.60 after chunk and retrieval fixes Work through the playbook and rebuild index parameters. Retrieval Playbook

• Answers flip between runs or sessions Verify version skew and session state. Pre-Deploy Collapse

---

🔗 Quick-Start Downloads (60 sec)

Tool	Link	3-Step Setup
WFGY 1.0 PDF	Engine Paper	1️⃣ Download · 2️⃣ Upload to your LLM · 3️⃣ Ask “Answer using WFGY + <your question>”
TXT OS (plain-text OS)	TXTOS.txt	1️⃣ Download · 2️⃣ Paste into any LLM chat · 3️⃣ Type “hello world” — OS boots instantly

---

Explore More

Module	Description	Link
WFGY Core	Canonical framework entry point	View
Problem Map	Diagnostic map and navigation hub	View
Tension Universe Experiments	MVP experiment field	View
Recognition	Where WFGY is referenced or adopted	View
AI Guide	Anti-hallucination reading protocol for tools	View

If this repository helps, starring it improves discovery for other builders.