Give a coding agent project knowledge it cannot reliably infer from the code, put that guidance where its harness will find it, and test whether it changes results on a representative task. A useful project guide explains the repository’s structure, verified commands, architectural boundaries, conventions, and completion checks—not generic advice that could apply to any codebase.
Start with a recurring problem, not a blank instruction file
Before writing guidance, identify a repeated failure: perhaps an agent adds a feature in the wrong package, bypasses an established interface, runs the wrong tests, or overlooks a security requirement. Choose a representative task that exposes the problem and define observable success criteria. For example, record which files should change, which project conventions matter, which checks should run, and what corrections the task currently requires.
As an Amazon Associate I earn from qualifying purchases.
This gives the guide a purpose. Microsoft’s VS Code guidance on configuring AI for a codebase recommends establishing a baseline with a representative task before adding project instructions. It also emphasizes that project guidance is most useful for decisions an agent cannot reliably infer from code alone.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Choose an instruction format the agent actually discovers
There is no single instruction filename that every coding agent, IDE, version, and session necessarily reads. Choose based on the harness your team uses, and verify its current support and discovery behavior before relying on a file.
#1 Best Overall
- [Ultra-Lightweight Ear Buds Designed for Small Ears] Each earbud weighs only 3.7g and features a compact, ergonomic in-ear design made especially for small ears. Secure, low-profile, and comfortable for workouts, all-day wear or overnight sleeping.The ultra-slim profile sits flush in your ear with no hard edges poking or rubbing, so you can even sleep on your side without discomfort.
- [Immersive Stereo Sound with TOZO OrigX Technology] TOZO OrigX tuning delivers clear vocals, balanced mids, and natural stereo sound for music, podcasts, and videos.
- [Long Battery Life for Daily Use] Get up to 7 hours of playtime on a single charge, with up to 32 hours total using the charging case—ideal for workdays, commuting, and extended listening sessions.
- [Bluetooth 5.3 & Stable Connection] Bluetooth 5.3 provides fast pairing, stable wireless performance, and reduced dropouts as you move around home or office.
- [Deep Bass with Clear Vocals] High-performance drivers produce punchy bass while keeping vocals clean and detailed for everyday listening.
| Documented harness | Project instruction format | Scope notes |
|---|---|---|
| OpenAI Codex | AGENTS.md |
OpenAI documents directory-scoped guidance; more deeply nested files can take precedence in conflicts. Confirm how the selected Codex setup resolves instructions. |
| GitHub Copilot in VS Code | .github/copilot-instructions.md or AGENTS.md |
VS Code also documents targeted .instructions.md files with applyTo patterns. |
| Anthropic Claude in VS Code | CLAUDE.md |
VS Code documents path-targeted rules using paths. |
The supported formats and targeted mechanisms above are described in the VS Code documentation. Support can depend on the chosen harness and session. If a team uses multiple agents, share a format where each supports it; otherwise keep separate instruction files aligned so they do not conflict.
What to put in the project guide
Keep root-level guidance focused on facts that are important across the repository. The guide should help an agent orient itself and make changes in the project’s actual way, rather than repeat general programming principles.
Rank #2
- 【IP68 Waterproof Multifunctinal Smart Watch】Torjalph fitness activity tracker smart watch is IP68 waterproof technology ,you can wear it when swimming,handwashing,it's rain-proof which means that you perform multiple exercise modes outside without worrying about sweat or rain corroding your fitness smart watch,and this watch equipped with a 230 mAh large capacity battery magnetic charging ,and this sports activity tracker watch can be used 5-7 days If fully charged.
- 【Larger 1.83'' Full Screen Watch & Multi-dial Selection】1.83 inch full HD color touch screen smart watch for men / women provides excellent 240*284 resolution image quality ,This Bluetooth smart watch offers a better touch experience and remarkable immersion and ,more than 200 watch faces for you to choose from App"Da Fit ", you can change the dial style according to your mood to wear keeping the world on your wrist
- 【Accurate Activity Tracker Smart Watch 】sports watch activity tracker uses sweat and dust resistant breathable silicone strap, our Fitness Tracker Smart Watch for Android iOS Phones supports 12 sports modes,walking,running,bicycle,swimming,badminton,basketball,football,rope skipping. This bluetooth calls smart watch for men women is lightweight and stylish, bringing more convenient and relaxing sports experience while also providing more protection.
- 【Your 24 Hour Personal Daily Health Monitor】Fitness watch for women men full touch screen IP67 waterproof bluetooth smartwatch for Android iOS Phones with heart rate blood pressure monitor multifunction Andriod smart watch with steps counter sleep monitor sports fitness watch, monitoring blood oxygen saturation allows you to understand your physical condition in real time to track fitness progress ,to know your health status in time.
- 【More Functions Bluetooth Call Smart Watch】This fitness activity smart watch full screen for iPhone and Android phones includes a bunch of new features including:The bluetooth smart watch can hang up the phone, receive and display WhatsAPP /Text /Facebook /Twitter /Messenger reminder.Support sedentary reminder, alarm clock, stopwatch, timer,weather information,music control, remote camera control compatible with Android 5.0 and above or iOS 9.0 and above smartphones.
Purpose and repository map
Describe what the project does and identify the major areas an agent needs to navigate: application code, tests, examples, documentation, generated files, and important packages. Explain relationships that are not obvious from names alone, such as which service owns a responsibility or where new endpoints belong.
Setup and verified commands
List the install, build, test, lint, and formatting commands that have been checked in the current repository. Include the working directory when it matters, and note when a command applies only to a package or change type. Do not provide a command just because it is customary; an incorrect command wastes time and can lead to false confidence about validation.
Rank #3
- [Low Noise High Fidelity] New bee guitar cord with 22AWG oxygen-free-copper center conductors provides enhanced signal transfer, the PVC insulation material and the high-density stranded copper double shield provide noise rejection for quiet operation
- [Ideal for Choice] New bee 1/4 straight and right angle metal connectors, ensures a secure connection in any instrument, pedal, or amp jack. L shape great fro on stage guitar stands
- [Wide Compatible] Suitable for mixer, power amplifier, microphone, effector, decoder, equalizer, electoronic organ, electric guitar, electric bass and other electronic equipments. Widely used in the installation of smart home, karaoke speaker, professional stage audio, etc
- [Easy Disassemble] Making incidental repairs quick and easy! Very nice, flexible cable
- [Dont Hesite to Order] 24 hours customer services and professional technology team are standing by
Architecture and boundaries
Explain where new functionality belongs, which interfaces should be used, and which dependencies or layers must not be bypassed. State relevant constraints on compatibility or data flow. Code can reveal what currently exists; it may not reveal why a boundary matters or what the project expects future changes to preserve.
Project-specific conventions and requirements
Include conventions that materially affect changes: naming, error handling, API compatibility, security requirements, and testing expectations. Prefer concrete directions tied to this repository over broad instructions such as “write clean code.” If a rule is already maintained authoritatively elsewhere, link or point to that source instead of creating a conflicting duplicate.
Rank #4
- Digital Family Calendar: BSIMB digital calendar wall touch screen puts an end to your chaotic life. Designed for busy families, the digital calendar keeps you organised and in step with your family. Assign different colours and tasks to each family member to make family scheduling easier and more intuitive.
- Easy Setup and Auto-Sync: Plug in, connect to WLAN and auto-syn your electric calendar in minutes. Digital wall calendar compatible with Google calendar, iCloud calendar, Outlook calendar, Cozi calendar and Yahoo calendar. Add events directly on the device or through the ecalendar app. Electronic calendar stays current whether you're at work or traveling—no manual updates needed.
- All-In-One Home Assistant: Smart calendar builds habits and encourage independence with personalised chores. Display daily meal plan to reduce the daily ‘What's for dinner?’ stress. To-do lists keep track of everything you want. Turning chores into valuable achievements with stars and rewards, making life easier and more enjoyable.
- Manage Anytime Anywhere: BSIMB electronic calendar for family stay connected with the powerful ecalendar app. Access and manage your digital wall calendar anytime, anywhere with free features like stars & rewards, photo screensaver, chore charts, meals plans, to-do lists etc. Electronic planner paid features like Magic Import tailor advanced tools for users with special needs.
- Stay in Touch with Loved Ones: Share photos or videos from your phone to electronic wall calendar with app, instantly update your life moment, and keep in touch with your family, especially those who are far away; you can also surprise your smart calendar digital wall by sending him a wish card; it's also great for keeping tech-illiterate elders in touch with the youngsters.
Completion checks
Define what “done” means for common change types. A documentation edit may need no build; a change to a particular package may require its tests and formatter. Make checks conditional where appropriate rather than giving every task an indiscriminate, lengthy checklist.
Free tools Windows power users keep installed
One-click scans. No signup required.
Use nested or targeted instructions for local rules
Not every convention belongs in the root guide. Put a rule closer to the code it governs when it applies only to a language, package, module, or set of files. This keeps global guidance shorter and helps avoid distracting agents with irrelevant constraints.
For Codex, OpenAI describes AGENTS.md instructions as scoped to the directory tree rooted at the file, with deeper files taking precedence in conflicts. VS Code also documents subfolder AGENTS.md files for Codex. In VS Code, targeted instruction files can use applyTo patterns, while Claude rules can use paths. These resolution details are harness-specific; check the documentation for the exact agent and session, especially when instruction files overlap.
Write and maintain instructions through a short validation cycle
- Record the baseline. Run the representative task without the new guidance under a defined harness and comparable settings. Note the changed files, conventions followed or missed, commands run, and corrections needed.
- Add the smallest useful rule. Write only project-specific guidance that addresses the observed failure. Keep generally applicable facts in the root guide and local rules in supported targeted or nested files.
- Review every statement. Check that directory names and paths exist, commands work from the stated location, and architectural or security requirements are accurate. Remove speculative, obsolete, generic, or duplicated text. VS Code’s configuration guidance likewise recommends reviewing the repository map, commands, technology choices, conventions, and definition of done rather than assuming generated instructions are correct.
- Confirm discovery in a fresh session. Use the same harness and repository, start a fresh session, and verify that the intended instruction file or targeted rule is visible to the agent. A file being present in the repository does not by itself establish that a particular tool loaded it.
- Repeat and compare the task. Keep the task and environment as similar as practical. Compare changed files, adherence to conventions, commands run, and the amount of correction required. Revise guidance only when the results show a gap.
How to tell whether the guidance helped
Separate two questions: did the harness discover the intended instructions, and did the agent follow them? Seeing an instruction file in the tool’s references or customization view answers the first question, not the second. The VS Code guide explicitly cautions that discovery does not prove adherence to every instruction.
Compare the baseline and updated runs using the success criteria you set in advance. Look for observable differences in repository choices, implementation boundaries, validation commands, and corrections needed. Keep the task and settings comparable enough for the difference to be informative. One successful run is evidence about that task and setup, not proof of a universal performance improvement. Official guidance reviewed here does not establish a general improvement percentage, so results should be reported in terms of what changed in your own project.
Recommended Free Tools
Quick Recap
Common mistakes to avoid
- Writing generic rules first: instructions that do not address a real repository failure add noise without a clear success measure.
- Assuming a filename is universal: format support and discovery vary by harness, configuration, and session.
- Putting every rule at the root: module-specific guidance can distract from the task and make global instructions harder to maintain.
- Trusting generated text without review: paths, commands, architecture descriptions, or completion checks may be inaccurate or stale.
- Duplicating a source of truth: repeated rules can drift or conflict; point to the authoritative project documentation where practical.
- Treating discovery as success: loading a file does not show that the agent understood or followed its contents.
- Claiming a general performance gain: evaluate against a comparable task and report project-specific observations rather than an unsupported percentage.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
