Official product page from **DeepSeek**, published on **August 13, 2026**, **unsigned**, ~450 words, announcing the *developer preview* release of **DeepSeek Harness** (`dsh`) — a coding-agent harness **open source under the MIT license**, whose repository opened the same day. A three-word thesis, repeated in the title and in the repository description: *« Everything is a plugin »*, paired with a second promise, *« Every run is traceable »*. The page states the equation *« AGENT = MODEL + HARNESS »* and lists the pluggable capabilities — *« models, tools, skills, sessions, sandboxes, storage, loops, scheduling, and the UI »*. Four modes ship: **Standard** (full coding agent), **Code** (tools exposed via the *Code Mode SDK*, letting the model compose multi-step operations inside a TypeScript program), **Minimal** (*« two-tool coding agent with persistent bash and str_replace_editor »*, explicitly *« for benchmarking models in a minimal environment »*), and **Creator** (runtime inspection, in-memory plugin testing). The technical substance sits in the repository, not on the page: `docs/architecture.md` states a logging invariant — *« Model-visible means logged. Anything that reaches a model request must be reconstructable from the log, and a runtime invariant asserts it »* — and states that *« there is no privileged core to patch »*. The technical core is not DeepSeek's own: DSH is built on **Cordis** (the `cordiverse` project, a third party), **vendored** into `vendor/` with a manifest and a sync procedure, and the page places the *« Cordis paper »* at the same navigation level as "GitHub" and "Developer docs". Two LLM adapters ship — `dsh-llm-deepseek` and `dsh-llm-pi-ai`, a generic multi-provider adapter. The repository warns in capitals: *« THERE WILL BE COMPATIBILITY-BREAKING CHANGES »*, and `CLAUDE.md` specifies that `SESSION_FORMAT_VERSION` stays at `0` *« with no compatibility promise »*, with backends rejecting old on-disk formats. Timeline: DSH ships on the day **DeepSeek-V4-Pro reaches GA**, three days before a new API pricing schedule takes effect on **August 16, 2026 at 16:00 UTC**, with peak/off-peak rates and an off-peak discount of **−50%**.
#DeepSeek Harness#dsh#agent harness
**DeepSeek** (DeepSeek AI, laboratoire chinois) · en tant qu'institution. Page produit **non signée** : aucun auteur · aucun ingénieur mis en avant · aucun billet de blog ni papier technique associé. Le « nous » n'apparaît qu'une fois · en dernière phrase — *« We look forward to exploring the limits of intelligence with developers worldwide »*. Publiée le **13 août 2026**. La page est rendue en JavaScript : `curl` sur l'URL renvoie **HTTP 202 avec un corps vide** · le texte n'existant qu'après exécution du bundle. Deux documents de politique sont liés en pied de page — *Safe Use Policy* et *Data Processing Statement*.
A watch note by **Didier Girard** published on **X** on **August 7, 2026**, which reads the launch of **Shieldstral 1.0 3B** (Mistral AI, August 4, 2026) not as a product release but as **the production deployment of a doctrine**. Starting point: on **May 13, 2026**, before the National Assembly's commission of inquiry into digital vulnerabilities, **Arthur Mensch** refused any oversight role for Mistral over the end use of its models — *"we do not have democratic legitimacy"* — explicitly rejecting **Anthropic**'s stance. Less than three months later, Mistral releases a **moderation model**. The author dismisses the apparent contradiction: **Shieldstral carries no taxonomy of the licit and the illicit**, it answers a **question the user writes**. **The mechanism is the heart of the note**: a three-part prompt (context + severity / a single closed question / the content to be judged), a `yes` or `no` response, and the **softmax over these two tokens** produces a continuous score between 0 and 1. **The moderation policy is not in the weights, it is read at inference time** — whereas **Llama Guard 4** embeds the MLCommons taxonomy fixed at training time, Shieldstral reads yours in natural language, modifiable **without retraining**. The technical report (**arXiv:2607.25857**, July 28, 2026) quantifies the cost of this choice: fine-tuning on public data alone = **61.1% F1** on policy adaptability; **4.4 million contrastive pairs** generated by an LLM (the same content rewritten to violate a policy but not its sibling policy) = **+23.3 points**; **91.3%** after merging three checkpoints. Characteristics: **3.8B actual parameters** (the "3B" in the name rounds down), **Ministral 3** base + **Pixtral** vision encoder, **12 languages**, **16 GB of VRAM in BF16**, **Apache 2.0**. Text performance: **84.9% average F1**, on par with **GPT-OSS-Safeguard-20B** (seven times larger), ahead of **Qwen3Guard-8B** (84.0) and far ahead of **LlamaGuard-4-12B** (69.1). **A caveat raised by the author himself**: *all these figures come from Mistral, on test sets selected by Mistral, and no third-party evaluation existed as of August 6*. The note's structuring thesis is an **opposition of topologies**: at **Anthropic**, the guardrail lives **in the weights** and the publisher arbitrates who is exempt from it (**Claude Fable 5** public with safety measures / **Claude Mythos 5** without, reserved for approved cyberdefenders of **Project Glasswing**, June 9, 2026); at **Mistral**, the guardrail **sits outside the model** — a separate, open, self-hostable component, whose policy belongs to the deployer. Explicit customer alignment (ministry of the Armed Forces, BNP Paribas, French and Luxembourg government administrations). The note closes on a **setback documented in three points**: **auditability** (binary output, no reasoning trace, while the deployer inherits the burden of justification under an AI Act audit), **robustness** (the first chapter of Voltaire's *Treatise on Tolerance* classified as "calls for violence" by a tester on the Hacker News thread — a mention/endorsement confusion), **availability** (as of August 6: no billed endpoint on La Plateforme, no official Ollama). Three deployment rules to close.
#Shieldstral#Shieldstral 1.0 3B#Mistral AI
**Didier Girard** — auteur de la note · publiée sur son compte X. Écrit ici en **analyste de doctrine industrielle** plutôt qu'en testeur : il n'a pas déployé le modèle · il croise une **audition parlementaire** (Mensch, 13 mai) · un **lancement produit** (Shieldstral, 4 août) · un **rapport technique** (arXiv, 28 juillet) et un **contre-exemple concurrent** (Anthropic, 9 juin) pour montrer qu'ils forment une position cohérente. Deux marqueurs de posture : il **borne explicitement la valeur des chiffres** qu'il cite (aucune évaluation tierce) et il **termine par des règles opérationnelles** — l'analyse doit sortir avec sa traduction en décisions de déploiement.
**Skill** entry: **hyperresearch** by **Jordan Gibbs** is a **deep research harness** that turns Claude Code into a documentary research agent, shipped as a PyPI package (MIT, Python 3.11-3.13) installing **20 Claude Code skills**, a CLI, an MCP server, and a local web UI. Observed on **August 3, 2026**: 1,568 stars, 170 forks, repo created on April 9, 2026, last push on August 1. **The core is a 16-step pipeline adaptive by tiers** — `light` (~30-40 min), `full` (~1.5-2.5 h), `dissertation` (4-8 h, 25,000-80,000 words across 300-450 sources) — which takes a prompt and returns an adversarially audited report with full provenance. **The central architecture decision is documented alongside its failure mode**: the entry skill is a **thin router** with no procedure, each step living in its own skill loaded **fresh at the moment it is invoked**, because the previous version was *« one 1200-line skill that got compacted away by the time Layer 4 needed its triple-draft procedure. The orchestrator forgot the procedure, wrote a single draft, and produced a flat-scoring report. »* **Two load-bearing principles.** *« Patch, never regenerate »*: after synthesis, only surgical `Edit` touch-ups are possible, with the patcher and the polish auditor tool-locked to `[Read, Edit]` at the Claude Code allowlist level, so that they *« physically cannot Write a new draft »*. *« Canonical research query is gospel »*: the verbatim prompt is persisted once in `query.md` and re-read by every step and every subagent. **Sixteen subagents** with configurable role and model (fetchers and cite-checker on Sonnet, critics, synthesizer, and patcher on Opus). **The vault** is a persistent markdown store indexed in SQLite — *« Markdown is truth, SQLite is cache »* — with a note lifecycle (`draft → review → evergreen`, `stale → deprecated → archive`), traceable provenance, a composite quality score (source type, citation authority via OpenAlex and Semantic Scholar with retraction flags, internal PageRank), and an **independence audit** that groups syndicated copies together — *« five reprints of one press release argue with the weight of one source »*. **Three mechanical gates before shipping**: citation integrity (every quoted citation must exist **verbatim** in a vault note), a retraction sweep refreshed on every cited DOI, and a citation-to-sentence link check by a skeptical LLM. **Reservation to flag**: the opening claim — *« currently leads the DeepResearch-Bench RACE leaderboard »* — is contradicted by its own footnote, *« forward-looking projection from a stratified pilot… Third party validation is pending »*. A projection is not a ranking, yet the chart places it ahead of Gemini and OpenAI Deep Research.
#skill#deep research#research harness
**Jordan Gibbs** — auteur et mainteneur du dépôt `jordan-gibbs/hyperresearch`. Le projet est distribué sous **licence MIT** et publié sur **PyPI** (`pip install hyperresearch`). Signaux d'adoption au 3 août 2026 : **1 568 étoiles** · **170 forks** · 13 issues ouvertes · dépôt créé le **9 avril 2026** et poussé le **1er août 2026** — soit une traction rapide sur moins de quatre mois. Topics déclarés : `agents` · `agentskills` · `claude-code` · `deep-research` · `deep-research-agent`.
SFEIR's decryption (firm voice) of Jason Clinton's (Deputy CISO, Anthropic) debrief published five days earlier — already documented in [[clinton-anthropic-secure-ai-native-sdlc-2026-07-21]]. **The added value lies not in the facts but in the thesis that rereads them**: if Anthropic's controls hold, it is because **a cycle with named stages exists to hang them on** — "the SDLC is the foundation, not a formality." The demonstration proceeds by rereading the mapping (**PSR at Plan, CLAUDE.md + egress allowlist at Code, review agents at Test, continuous DAST at Deploy, triage + SIEM routing at Monitor**), then through a **four-part anaphora**: (1) *without an SDLC, productivity gains do not materialize* — Clinton cites **Amdahl's law**: multiplying code volume by 8 multiplies nothing if review stays sequential and human, and Anthropic gained not by distributing agents but by **identifying the blocking stage (Test) and rebuilding it** — "you don't optimize a bottleneck you haven't mapped" (echoing DORA 2025's **mirror effect**); (2) *without an SDLC, security has no anchor point* — a **gate is by definition a control placed between two stages**, and Clinton's three threats are addressed at distinct moments; (3) *without an SDLC, no **token FinOps** policy can be formulated* — agentic scanning is billed on consumption and grows with code throughput, so **risk-based tiering IS the FinOps policy** (it decides where three agent passes get paid for and where a SAST suffices), otherwise "token spend is not steered, it is discovered at month's end"; (4) *without an SDLC, there is nothing to measure* — the indicators (16% → 54% of PRs commented, one third of past incidents intercepted) exist only because there are stages where a counter can be placed; absent that, one produces only **usage figures** (licenses, tokens) that say nothing about quality or risk. Two strong points beyond the thesis: the reading of the **incident agent-à-agent** ("a security perimeter that rests on an instruction in a prompt is not a perimeter"; **an agent's access to other agents is part of its attack surface**) and an **explicit methodological caveat** — Anthropic's figures about Anthropic, unaudited, published by the vendor of the model described, in the context of a young codebase with no mainframe: **what transposes is the method, not the figures**.
#SDLC#AI-native SDLC#development cycle
SFEIR (voix éditoriale du cabinet, article non signé individuellement) — commentaire de Jason Clinton (Deputy CISO, Anthropic)