# freecode.academy

**Website:** https://freecode.academy

## What It Is

Freecode Academy is positioned as an **AI R&D Lab + private execution network**, rather than a classic freelance marketplace. The core idea is not just to connect clients and contractors, but to participate in the full cycle: clarify the problem, assemble the right team, organize collaborative execution, and maintain responsibility until the actual delivery of results.

The network has existed since 2012 and was reimagined for the AI era in 2026. The website states **4,000+ members** and **100+ studios and teams**.

## Key Principles

* **Curated / invite-only:** restricted access for high signal-to-noise and trust among participants.
* **AI-native:** AI is used not only as a production tool, but also as a navigation and matching layer for people, projects, and offers.
* **Shared responsibility:** the platform does not end its involvement after introducing the parties, but declares shared responsibility for the result.
* **Real collaboration:** emphasis on collaborative teams and real projects instead of random one-off gig orders.
* **Research → execution:** public research and experiments should transition into prototypes, products, and working systems.

## Main Working Cycle

1. **Define the problem** — turn a vague idea into a clear scope.
2. **Assemble the team** — select specialists and teams within the verified network.
3. **Execute together** — work in a coordinated team with shared goals and transparency.
4. **Deliver results** — drive the work to a concrete outcome.

## Main Platform Entities

* **Members** — network participants; public viewing is open, direct contact is restricted to invited participants.
* **Teams** — studios and working teams that can accumulate portfolios and jointly take on more complex projects.
* **Offers** — a flexible format for proposals: service, collaboration, research, hiring, and other initiatives.
* **Projects** — projects with the opportunity to search for partners, investments, and assistance.
* **Tasks** — specific bounded tasks within projects; completed work can become a public worklog/case study.
* **Topics / Publications** — the research and public layer of the platform.

## Who It Is For

### Clients

For companies and founders who need not just to find a contractor, but to navigate the AI chaos, determine the right solution, and assemble a reliable team for implementation.

### Specialists

For professionals who need real projects, strong teams, reputation, and access to a closed deal flow.

### Studios and Team Leads

For teams that want to combine competencies, scale up to larger tasks, and collaborate with other studios.

## Historical Context

Freecode.academy retains an earlier educational layer: learning sections on HTML/CSS/JavaScript/React, algorithms, data structures, and practical projects. In past versions of the project, much attention was paid to teaching web development through real public tasks and pull request workflows. The current positioning is significantly broader: a closed professional network and AI-native project execution infrastructure.

## Why the Project Matters

freecode.academy combines several areas that usually exist separately: a professional community, R&D, project management, team matching, reputation, and AI-assisted matching. The main product hypothesis is that **network quality and responsibility for results are more important than the open scale of a marketplace**.

## Research Sources

* https://freecode.academy/
* https://freecode.academy/about
* https://freecode.academy/projects
* https://freecode.academy/start/developers

Freecode Academy is a private AI-native execution network: a platform that brings together clients, independent specialists, and studios, helping to formalize tasks, assemble teams, and drive projects to results with shared responsibility.

freecode.academy

Need to deploy the fresh version of the heych agent.

Update website

# Problem

On `freecode.academy`, some URLs with percent-encoded special characters may return a 404, even though the corresponding page likely exists.

Example:

```text
/comments/topics/dzheneriki-typescript/nu-vot-ts-tebe-govorit-%22ty-zanimaeshsya-figney%22,-i
```

Hypothesis: at one of the routing stages, slug lookup, or URL matching steps, URL decoding (`decodeURIComponent` or equivalent) is either missing or incorrectly performed, causing the slug to be compared in its encoded form against the stored decoded value, or vice versa.

## What to Check

1. Trace the full path of the URL from the HTTP request to the entity lookup by slug:
   - framework/router;
   - route params;
   - server/API handler;
   - DB query;
   - canonical URL/link generation.
2. Check at which stage `%22` and other percent-encoded characters are decoded automatically, and where explicit decoding is required.
3. Prevent double decoding (`decodeURIComponent` over an already decoded param), especially for `%`, `%25`, `%2F`, and other sensitive sequences.
4. Verify the slug format in the DB: whether it is stored decoded, encoded, or normalized.
5. Check the link generator: URLs must be encoded exactly once (`encodeURIComponent`/URL API where necessary).
6. Specifically test the following characters:
   - quotes `"` / `%22`;
   - commas;
   - spaces / `%20`;
   - Cyrillic characters;
   - `%`;
   - `?`, `#`, `/` inside a potential slug.
7. Check old, already published URLs and backward compatibility. If different historical variants of the same slug exist, add a redirect/canonical normalization instead of a 404 where necessary.

## Specific Case

For the URL:

```text
/comments/topics/dzheneriki-typescript/nu-vot-ts-tebe-govorit-%22ty-zanimaeshsya-figney%22,-i
```

we need to find out:

- whether a record with the logically corresponding slug exists;
- what value the application actually receives as the route param;
- what value is used in the DB query;
- whether the 404 is fixed after proper normalization/decode.

## Acceptance Criteria

- [ ] The exact cause of the 404 for the given URL has been found.
- [ ] URLs with special characters correctly open existing pages.
- [ ] There is no double decode and related errors/exceptions.
- [ ] Link generation and route param reading use a consistent slug format.
- [ ] Invalid URLs still return a proper 404 rather than causing an application error.
- [ ] Tests have been added for at least `%22`, Cyrillic characters, and `%25`/double encoding.
- [ ] Backward compatibility of existing links has been verified; a redirect to the canonical URL has been added if necessary.

Check 404s for URLs with special characters and the correctness of decode/slug routing.

Tasks

Double-check URL handling and special character decoding

Update website