The OIAR Skillblog
Every article is an OIAR skill written for people: the knowledge the agents use to interview, analyse and write the report. You read it here, you use it in OIAR.
What is a Skillblog?
Every OIAR framework has a Skill Loader: a loader of skills that the agents pick according to the context, organised by type. A skill can be about a product, a methodology or a vertical process.
This is what the blog is for: every skill of the catalogue has its article here, in plain language, so you know what the agent knows before you hand it a project. The version for machines stays in OIAR, where it works.
Designing stable APIs: contract first, consistent errors, add instead of change
Principles for interfaces that last: Hyrum's law, a single version, contract first, uniform errors, validation at the boundaries, additive changes, predictable names. Patterns for REST and types.
Read the article →Testing in a real browser: console, network, DOM and performance with developer tools
Verifying what runs in a browser with developer tools: DOM, console, network, performance, accessibility, test plans and untrusted content.
Read the article →Code review on several axes: correctness, readability, architecture, security, performance
How to review a change before it enters the main branch: five axes of assessment, size of changes, severity of findings, and how to review an agent's work.
Read the article →CI/CD: automatic quality gates and releases you can repeat
How to set up integration and release pipelines: automatic checks on every change, tests in the pipeline, secrets, gradual release strategies and rollback.
Read the article →Simplifying code without changing its behaviour
How to make code that works but is hard to read clearer: Chesterton's fence, the rule of three, small steps with green tests, and when not to touch.
Read the article →Systematic debugging: stop the line, reproduce, locate, fix the cause
The six-step method for finding the cause of an error instead of trying things at random: reproduce, locate, reduce, fix at the root, guard against recurrence, verify.
Read the article →Deprecating and migrating: switching off an old system without leaving anyone behind
How to decide whether to keep or retire a system, and how to migrate users: code as a liability, deprecation with a deadline, gradual migration, removal only once usage is zero.
Read the article →Documenting decisions: ADRs and documentation for whoever comes later
How to record the why of technical choices: architectural decisions (ADRs), documentation of APIs and features, comments that explain the reason. For whoever will read it in a year, person or agent.
Read the article →Doubt-driven development: have every important decision attacked by a reviewer with fresh context
A five-step method for testing the decisions that matter: state the thesis, extract the minimum unit, ask for an adversarial analysis with fresh context, reconcile, stop.
Read the article →Quality user interfaces: components, state, accessibility and well-made empty states
How to build an interface that does not look generated: component architecture, state management, consistency with the design system, accessibility, responsiveness, loading and error states.
Read the article →Git as a discipline: atomic commits, short-lived branches and a history you can read
How to use git so that the project's history really helps: small, descriptive commits, short-lived branches, conflicts resolved calmly, tagged versions, parallel work.
Read the article →Refining an idea: from five variants to a direction with a list of things you will not do
A three-phase method for turning a vague idea into a concept to build: expand with a few reasoned variants, converge on one direction, state assumptions and what will not be done.
Read the article →Incremental implementation: small steps you can try and undo
How to deliver a change in increments: vertical slices, one thing at a time, code that always builds, features behind a switch, safe defaults, easy rollback.
Read the article →Optimising performance: measure first, fix the real bottleneck
How to improve performance without guessing: measurable goals, profiling, the bottleneck, the fix and a new measurement. Core Web Vitals, queries, payload size.
Read the article →Planning and breaking down work: vertical slices, small tasks, checkpoints
How to go from a specification to an executable plan: dependency graph, vertical slices, tasks with acceptance criteria, sizes, order and checkpoints between phases.
Read the article →Code security: untrusted input, authentication, data and integrations
How to make code that accepts untrusted input, handles sessions, stores data or talks to external services resistant: boundaries, authorisation, secrets, dependencies, logs.
Read the article →Specification before code: a living document with three-level boundaries
Why write a specification before building: goal, commands, structure, tests, three-level boundaries and a phased flow with a check for each phase.
Read the article →Data migration in a utility: from the old CRM and billing to the new one
Phases, quality KPIs, reconciliations and handling of rejected records to move customers, contracts and invoices without losing trust in the data.
Read the article →Context engineering: giving an agent the right context, in the right place
How to prepare an agent's context: rules files, specifications, relevant files, errors, managing the conversation. Packaging strategies and how to treat conflicts.
Read the article →Getting ready to ship: checklist, monitoring, gradual rollout and a rollback plan
What to check before going to production: quality, security, performance, monitoring, gradual rollout, rolling back, and the first hours after release.
Read the article →Source-driven development: every framework choice rests on the official documentation
How to avoid outdated patterns: detect the stack and versions, read the official documentation, follow documented patterns and cite the sources. The signs that you are working from memory.
Read the article →Test-driven development: red, green, refactor, and the proof of bugs
The TDD cycle, the bug-proof pattern, the test pyramid, how to write tests that describe behaviour, and the most common anti-patterns.
Read the article →Change requests and starting application management: governing change before and after go-live
A lean process for change requests and the handover to the AMS service in a CRM/billing project for an energy retailer.
Read the article →Odoo Manufacturing: bill of materials, work centres and orders that come from demand
What Odoo Manufacturing 20 covers, how it is configured and what to check: bill of materials, orders, work centres, subcontracting, MPS, scrap and when custom development is needed.
Read the article →Odoo Marketing: dynamic lists, automation, deliverability and consent
What Odoo Marketing 20 covers, how it is configured and what to check: dynamic lists, automation, deliverability, double opt-in, events, SMS and when custom development is needed.
Read the article →Odoo Planning: roles, batch publishing and open shifts
What Odoo Planning 20 covers, how it is configured and what to check: roles, shifts, automatic scheduling, publishing, field service and when custom development is needed.
Read the article →Odoo Point of Sale: sessions, payment methods and cash closing
What Odoo Point of Sale 20 covers, how it is configured and what to check: sessions, payments, employees, loyalty, restaurants and when custom development is needed.
Read the article →Odoo Project: hours logged every day, billing from timesheets, budget under control
What Odoo Project and Timesheets 20 cover, how they are configured and what to check: stages, hours, time-based billing, budget, profitability and when custom development is needed.
Read the article →Odoo Purchase: from the demand signal to three-way matching
What the Odoo 20 Purchase module covers, how to configure it and what to check: demand, vendor pricelists, three-way matching, blanket orders, returns and when custom work is needed.
Read the article →Odoo HR: contract, work entries and payslips without end-of-month surprises
What Odoo 20 HR covers, how to configure it and what to check: contract, work entries, time off, attendances, payslips, recruitment and when custom work is needed.
Read the article →Odoo Inventory: one, two or three steps, routes, reordering and valuation
What the Odoo 20 Inventory module covers, how to configure it and what to check: product, flows, locations, routes, reordering, valuation, adjustments and when custom work is needed.
Read the article →Odoo Sales and CRM: an observable pipeline, quotations in Odoo, price lists as discipline
How to configure Odoo Enterprise CRM and Sales: pipeline stages, mandatory activities, the quotation-order-delivery-invoice flow, invoicing policy and price lists.
Read the article →Odoo Accounting: source document first, automatic taxes and locked closings
What the Odoo 20 Accounting module covers, how to configure it and what to check: chart of accounts, journals, taxes, reconciliation, lock dates, Italian e-invoicing and when custom work is needed.
Read the article →Odoo Helpdesk: teams, SLAs and tickets linked to returns and repairs
What Odoo 20 Helpdesk covers, how to configure it and what to check: teams, stages, SLAs, ratings, Help Center, after-sales and when custom work is needed.
Read the article →Odoo Website and eCommerce: a single catalogue, tested payments and configured SEO
How to set up Odoo Enterprise Website and eCommerce: a catalogue shared with the warehouse, payment providers, SEO per page, B2B and B2C, live chat connected to CRM and Helpdesk.
Read the article →Agile delivery assessment: telling ceremonies from results
How to assess an agile team: backlog, velocity, definition of done, technical debt, impediments. The questions by role, the evidence and how to recognise an adoption that is only a facade.
Read the article →Cloud assessment: what to check in an AWS, Azure or GCP environment
Architecture, identity, backup, CI/CD, observability and costs: the questions by role, the evidence and the warning signs of an assessment of a cloud environment.
Read the article →GDPR assessment: how to really check the processing of personal data
Record of processing, legal bases, DPIA, data subject rights, data breach, external processors. The questions by role, the evidence and the warning signs of a privacy assessment.
Read the article →Assessment of a SAP system: what to look at before deciding on S/4HANA
How an existing SAP (ECC or S/4HANA) is assessed: custom code, transports, authorisations, licences. The questions by role, the evidence to ask for, the warning signs.
Read the article →Network security assessment: segmentation, remote access and detection
What to look at in a corporate network: segmentation, firewall rules, remote access, SIEM, penetration testing. Questions, evidence and warning signs of a network assessment.
Read the article →