PRAXIS is a framework that enables agents to systematically extract, represent, and reuse tacit knowledge for domain code generation and outperforms state-of-the-art agents equipped with powerful agentic search capabilities, as well as experience-based and skill-based methods.
Abstract
LLM agents have achieved strong performance on general software engineering tasks, yet struggle with domain-specific code generation. We identify the root cause as the agent's lack of tacit knowledge, including domain-specific business rules, interface contracts, and operational conventions that developers internalize through practice but never document. This knowledge is deeply buried beneath the domain code, dispersed across code entities and their dependency relations, and invisible to the agent that lacks it. These properties make tacit knowledge inherently difficult to retrieve or learn. In this work, we propose PRAXIS, a framework that enables agents to systematically extract, represent, and reuse tacit knowledge for domain code generation. PRAXIS acquires tacit knowledge by simulating human development workflows within the target codebase, distills it into structured units organized on the code dependency graph, and proactively surfaces it to the agent at the point of code interaction. Extensive experiments demonstrate that PRAXIS outperforms state-of-the-art agents equipped with powerful agentic search capabilities, as well as experience-based and skill-based methods. The approach integrates seamlessly into various agent frameworks and LLMs with consistent performance improvements, and supports continual evolution with performance steadily scaling as practice accumulates.
AI coding agents need more than relevant snippets: they need business semantics, validation evidence, relations, and assurance that their context is current. Existing systems usually infer or externalize this knowledge through retrieval, summaries, graphs, rules, or reverse specifications. We investigate a complementary representation in which selected code units directly carry agent-usable knowledge. We introduce Executable Code Knowledge (ECK) and define an Executable Code Knowledge Unit (ECKU) as a source-bound object combining stable identity, semantics, executable behavior, contracts, evidence, relations, provenance, validation state, and a query interface. Our Python prototype supports code-local authoring, manifest export, evidence execution, exact changed-line impact, freshness checking, and agent-facing projections. Across three real Python repositories and 26 controlled patch tasks, direct ECK provides executable test coverage for 11/11 evidence-bearing tasks and exact selectors for 9/11; hiding declared evidence reduces exact recovery to 1/11 (paired exact McNemar p=0.0078). ECK-derived rules recover 11/11 exact selectors, showing that rules are effective delivery artifacts while ECK supplies source binding, validation state, impact, and freshness. Exact changed-line impact matches independently authored labels on all 26 patches (12 unit links; precision, recall, and F1 all 1.000). AST-bounded fingerprints classify 50 positive changes and 17 unrelated same-file controls correctly, whereas static rules snapshots detect none of the 50 stale cases. Model-backed patch-review and cross-layer studies measure projection fidelity rather than independent impact discovery. These results support a hybrid architecture: retrieval for coverage, ECK for source and evidence governance, and projections for delivery.
Recent advances in Large Language Model (LLM) “agent” systems have moved language models beyond single-turn generation toward goal-directed workflows that can plan, ask clarifying questions, and iteratively refine outputs. In parallel, retrieval-augmented generation (RAG) has become a practical way to ground these agents in enterprise knowledge—enabling models to leverage internal documentation and policies without expensive and recurring retraining. However, RAG is only as strong as the underlying corpus: when key information is missing, outdated, or fragmented, retrieval cannot fill the gaps. In many organizations, assembling high-quality, up-to-date documents remains a persistent bottleneck, limiting both the reliability of downstream applications and the speed at which knowledge can be operationalized. To address this, we introduce the Assistant-Scribe-Knowledge Checker (ASK) framework for generating structured specification documents through guided interviews. ASK decomposes the end-to-end workflow into three specialized LLM agents: an Assistant that asks questions to the user and steers the interview through adaptive follow-ups; a Scribe that continuously summarizes what has been said and records it into a schema-constrained specification document with explicit content requirements; and a Knowledge Checker that evaluates the evolving document against those specifications, detects gaps or weakly supported entries, and advises the Assistant on the next best questions to ask to reach the desired completeness and quality. This separation supports seamless document creation while improving quality control over content, structure, and coverage. By distributing responsibilities across three lightweight, role-specific agents, ASK reduces reliance on a single large model and enables modular scaling across teams and domains. We evaluated the ASK framework in a consulting-firm setting where consultants are required to produce project “return-of-experience” documents (successes, failures, challenges, and solutions) to capture reusable knowledge. Compared to documents authored manually, ASK-guided interviews consistently produced more complete, better-structured, and higher-quality specifications with less variability across authors. Beyond measurable quality gains, consultants also reported a clear preference for the interview-based workflow, citing lower effort and a more natural way to articulate tacit project knowledge.
Sylvain Roudiere, Bianca Lento· European Conference on Knowl...· 0 citations
This work proposes TraceDev, a multi-agent framework for automated software development grounded in use cases that contain multiple functional points and complex semantics, and demonstrates the effectiveness of TraceDev in repository-level code generation from requirements.
Mingyu Chen, Yakun Zhang, Zihao Xie et al.· 0 citations
Safety-critical software systems in the automotive, aerospace, and industrial sectors have grown to millions of lines of source code, making traditional approaches to software documentation, architecture comprehension, and knowledge transfer increasingly impractical. As continuous integration and rapid release cycles accelerate software evolution, organizations face persistent challenges related to technical drift, loss of domain expertise, fragmented architectural knowledge, and the dependence on a small number of senior developers. Conventional code discovery techniques based on lexical or keyword searches provide limited support because they identify only literal text rather than the underlying functional semantics of software artifacts. This paper proposes the Dynamic Code Wiki, an intelligent software knowledge architecture that automatically transforms large codebases into continuously evolving, semantically searchable knowledge repositories. The proposed framework integrates Abstract Syntax Tree (AST) parsing, vector-based semantic embeddings, retrieval-augmented generation (RAG), and automated knowledge graph construction to generate structured documentation directly from source code and development artifacts. The architecture further strengthens compliance with functional safety standards by automatically linking software requirements to implementation components and performing call-graph analysis to identify subsystems affected by code modifications. In addition, engineering knowledge embedded within commit messages, code review discussions, defect reports, and software evolution history is preserved as a permanent, searchable organizational knowledge graph independent of individual developers. By combining semantic code understanding with automated documentation and traceability, the proposed framework significantly improves software maintainability, architectural transparency, impact analysis, and long-term knowledge preservation. The Dynamic Code Wiki provides a scalable engineering solution for modern safety-critical software development, enabling organizations to reduce knowledge loss, accelerate developer onboarding, strengthen regulatory compliance, and support continuous software evolution without relying solely on manually maintained documentation or the institutional memory of experienced engineers
Parth Govind Vanparia· International Journal of Eng...· 0 citations
This work proposes CURATE - Composition, User-in-the-loop, Reuse, and Automated Task Execution - a novel human-in-the-loop multi-agent system that uses LLM agents to manage and develop composable workflows across their entire lifecycle.
Nolan Cutler, Chia-Chen Kuo, Nanda Velugoti et al.· 0 citations
Legacy software repositories embed decades of domain knowledge in undocumented code, making understanding and modernization difficult. We treat a program as the implementation of an unobserved, declarative description of its computation and investigate whether making this latent declarative representation explicit improves repository-scale porting. ADFD-Migrate approximates the latent representation with an annotated data-flow diagram (ADFD) of processes, data stores, external entities, flows, and behavioral contracts. An LLM infers the source ADFD from bounded repository context, guided by static-analysis coverage checks. Dependency-aware chunking orders bounded process groups for target-language generation. Differences between the source ADFD and a statically recovered target ADFD then guide regeneration. We evaluate ADFD-Migrate on f2x50, a new benchmark of 50~Fortran repositories spanning 1.5k--1.6M lines of code and three complexity tiers, and assess the resulting ports along two dimensions: porting soundness, measured by source-oracle behavioral agreement, and porting completeness, measured by a composite migration outcome index. Against 382 curated Fortran-oracle probes, the generated Python passes 327 (85.6\%), with 40 repositories passing every attempted probe. ADFD-Migrate exposes all 382 planned behaviors as runnable targets, compared with 99 and 98 for direct and repository-context translation and 69 and 30 for the static-profile and dependency-chunking ablations. It also achieves a 93.1\% mean migration outcome index and a 17--59 percentage-point outcome-index advantage over direct translation on 47 repositories. These results suggest that an inspectable semantic bottleneck can improve the coverage and integration of repository-scale migration while enabling lower-cost generation for many repositories.
Shraddha Surana, Ashwin Srinivasan, Michael Bain· 0 citations