SShortSingh.
0
ProgrammingDEV Community ·

TypeScript Types Don't Validate API Request Bodies at Runtime

A developer explains that TypeScript types are erased at runtime, so they cannot guarantee the shape of incoming API request bodies. Using a type assertion simply tells the compiler to trust the data, which can let malformed or malicious input through. The article describes concrete risks, including server crashes from unexpected values, corrupted stored data, and mass assignment that lets clients set privileged fields like role. The recommended fix is runtime schema validation with a library such as Zod, which rejects bad input with a 400 response and strips unknown fields. Treating every value crossing a program's boundary as untrusted is presented as the core principle.

0
ProgrammingDEV Community ·

Meta issues $15 billion bond as investor demand tops $58 billion

Meta has reportedly raised $15 billion through a dollar-denominated bond sale, according to Bloomberg, drawing orders totaling around $58 billion. Investors are interpreting the strong demand as confidence in the long-term direction of Meta's VR and AR strategy, including its Quest headset ecosystem. The article's author, a developer of a vision-therapy app for amblyopia, argues that such financial backing supports stable hardware, operating-system performance, and developer APIs that specialized health applications depend on. The piece points to technical needs such as consistent high frame rates and improving tracking and passthrough quality. The author also promotes their own app, which is available in the Meta Quest store.

0
ProgrammingDEV Community ·

Python script checks XLSX metadata cleanup without altering the workbook

An article on DEV Community explains how to verify that an XLSX metadata cleanup changed only the intended properties. The author describes comparing an original demo workbook with a cleaned copy to confirm that author, company, and custom fields were handled as expected. The method treats XLSX files as ZIP packages and inspects the docProps XML parts. A standard-library Python script compares SHA-256 hashes of the other package parts to detect unexpected changes. The article notes that AI helped draft the text, while the code and results were checked against the demo files.

0
ProgrammingDEV Community ·

A Claude Code workflow built from past losses: five-layer harness guards against context and command failures

The author rarely encounters Claude Code's auto-compact warning because the workflow is structured to keep sessions well under half of the context window, since compaction replaces conversations with lossy summaries. The approach, termed harness engineering, treats the environment around the model (context, permissions, feedback, orchestration, and self-tests) as more decisive than the prompt itself. The harness was developed on a household pantry and nutrition tracker and grew through real incidents, starting with hooks added on 14 August 2026. Key components include a session-start hook that measures repository state directly and a shell-command guard that refuses irreversible actions while warning about misleading ones. The author also notes that piping test output through tail can mask failing exit codes.

0
ProgrammingDEV Community ·

Checking Test Results in Coding Agent Logs Is Harder Than Reading the Exit Code

A developer analyzed about 16,000 public coding agent runs to see whether the final test result could be determined by reading the last test command's exit code. The approach proved unreliable: for one model, 88 runs appeared to end on a failing test, but after corrections only 59 did, since many apparent failures were actually timeouts, signals, or non-test errors. Piped commands can mask the real exit status, and some codes such as pytest's 2 through 5 signal interruptions or the absence of tests rather than failures. The author recommends parsing the test summary line, stripping color codes, and ignoring coverage-gate messages and zero counts, while keeping an 'unknown' verdict for runs where no tests executed. Log readers also need to account for commands like git stash that can make a passing agent run look like a failure.

0
ProgrammingDEV Community ·

osFoundry, Codex 핸드오프 제안과 run 요청을 별도 툴로 분리한 이유

osFoundry의 MCP 서버는 로그인 전에도 initialize와 tools/list에는 응답해 클라이언트가 툴 목록을 볼 수 있게 한다. 목록에는 세션 작업을 요청하는 두 툴, propose_session_handoff와 request_session_run이 있으며 둘 다 세션이 일을 하도록 만든다. 핸드오프는 세션 오너가 Accept를 눌러야 큐에 들어가고, run은 Runner 자리를 준 시점에 이미 승인된 것으로 취급된다. 두 툴은 OAuth 스코프, 세션 내 역할, openWorldHint 어노테이션, 입력 필드와 길이 제한이 서로 다르게 설계되어 있다. 개발자는 이 차이를 Python 클라이언트로 재현해 라이브 스키마와 대조했다.

0
ProgrammingHacker News ·

Developer argues AI agent memory needs a standard format like OpenTelemetry

A post on the Hacker News forum argues that memory systems for AI agents need a shared, open standard, comparable to what OpenTelemetry provided for observability. The linked article, published on the LoreKit blog, makes the case for an open agent memory format. The post has gained limited traction so far, with 3 points and no comments at the time of reporting. The source material is brief, so the specific technical proposal and its details are not described here.

0
IndiaTimes of India ·

Massachusetts Couple Awarded $530,000 Over Dust and Noise From Nearby Mine

Josephine and Keith Beadling won a legal case against G. Lopes Construction Inc. and Bette Maki over a nearby mining operation. They had complained for years about dust and noise coming from the site. A jury awarded the couple $530,000 for pain, suffering, and property damage. The case highlights the environmental concerns that can arise when sand and gravel extraction takes place near homes.

0
IndiaTimes of India ·

How tomato extract pills were marketed as medicine in 1830s America

In the 1830s, Dr John Cook Bennett proposed that tomatoes could help with digestive problems. Following his claims, several entrepreneurs began producing and selling tomato extract pills as remedies. Rival manufacturers Archibald Miles and Dr Guy R. Phelps became embroiled in a public feud dubbed the Tomato Pill War. Newspapers across the country covered the dispute extensively, drawing considerable public interest.

0
IndiaNDTV ·

TDP-Led NDA Wins Andhra Rural Polls; Jagan Reddy Alleges Irregularities

The TDP-led NDA alliance has won the rural elections in Andhra Pradesh. TDP state president Palla Srinivasa Rao called the result a massive public endorsement of the NDA government's welfare and development agenda. Former chief minister Jagan Reddy has alleged irregularities in the polling process. The article provides limited further detail on the specific claims or the scale of the victory.

0
IndiaNDTV ·

Oil India Suspends Four Officials After Fire Death Toll Rises to Three

Oil India Limited has suspended four officials in Assam following a fire at one of its facilities. The death toll from the incident rose to three on Saturday after a critically injured worker succumbed to burn injuries. The suspensions were announced as the company faced mounting pressure over the fatalities. Full details of the incident are not available in the source excerpt provided.

0
IndiaTimes of India ·

BJP chief urges opposition to follow Advani's lead on national security

BJP president Nitin Nabin spoke at a special event marking L K Advani's 100th birthday. He called on opposition parties to unite when the country faces national security challenges, pointing to Advani's example. Nabin credited Advani and Atal Bihari Vajpayee with steering the BJP from its early years to its current prominence. The event also acknowledged Advani's contributions to the Ram Janmabhoomi movement.

0
IndiaIndian Express ·

Raipur man lured to Russia for work ends up jailed; family seeks help from CM and police

A man from Raipur travelled to Russia in search of employment but was instead imprisoned there. His family has since written to the Chief Minister and local police seeking assistance to secure his release. The article reports that the relatives are appealing for official intervention in the matter. Further details about the charges and the current status of his case are not provided in the source text supplied here.

0
IndiaNDTV ·

PM Modi Says Naxals Cleared From Forests, Warns of 'Dimagi Naxals' Among Public

Prime Minister Narendra Modi, speaking in Bastar on Diwali, said Maoist insurgents have been eliminated from the forests. He used the phrase 'dimagi Naxals' to describe those he claims remain hidden among the population. The remarks referred to recent agitations, which he alleged were backed by elements seeking to create chaos. The article provides limited further detail beyond these statements.

0
ProgrammingDEV Community ·

Integrated Knowledge Representation Project Outlines a Reusable Logic Inference Engine

This chapter from the STATERA Library's Volume 02 on Knowledge Representation presents an integrated project for building a logic inference engine. The engine derives new knowledge from explicitly stated facts and logical rules, aiming to turn formal reasoning principles into executable operations for symbolic AI. Its architecture comprises a knowledge base, a logical expression parser, an inference controller, working memory, and a query interface, with an emphasis on correctness, transparency, modularity, and extensibility. The material covers forward and backward chaining, unification, and resolution-based refutation as core reasoning mechanisms.

0
ProgrammingDEV Community ·

Neuro-Symbolic AI Combines Neural Learning with Symbolic Reasoning

Neuro-symbolic artificial intelligence is an approach that integrates neural network learning with symbolic knowledge representation and logical reasoning. It aims to pair neural networks' ability to recognize patterns in complex data with symbolic systems' capacity to express concepts, relationships, constraints, and rules. The approach emerged from the historical split between symbolic AI, which emphasized explicit logic, and connectionist AI, which focused on learning distributed representations from data. Integration is challenging mainly because neural outputs are continuous while symbolic systems operate on discrete entities, so uncertainty management and semantic alignment are needed. Strategies include using neural models to generate symbolic facts for reasoning engines, and embedding symbolic constraints into training, with differentiable reasoning and knowledge graphs playing key roles.

← NewerPage 454 of 7349Older →