0 Fork 0

mirror of https://github.com/paperclipai/paperclip-docs.git synced 2026-11-09 17:41:56 -10:00

Files

T

Aron PrinsandClaude Opus 5 325c9eb819 seo: hand-write titles and descriptions for all 192 pages (#115 )

## Why

The build derived `<title>` from the **sidebar label** and the
description from the **first paragraph clipped to 220 characters**. Both
are the wrong source. A sidebar label is written for a 240px-wide nav
that supplies its own context; a search result has none.

On production that produced:

- **7 pages titled "Overview"**, and 32 pages sharing a title with
another page
- **109 of 193 descriptions cut off mid-word**, and 174 over 160
characters (where Google truncates)

```
/guides/welcome/key-concepts/
  BEFORE (220):  ...and if you misunderstand them early, the product wi
```

Near-identical titles are the weak-differentiation signal behind
"crawled — currently not indexed", which is the exact symptom on this
property.

## What changed

`seo_title` and `seo_description` frontmatter on **every one of the 192
pages**, written by hand against each page's real content and optimised
for the click — lead with what the reader can do, name the concrete
things covered, no keyword stuffing, no repeating the title.

| | Before | After |
|---|---|---|
| Duplicate titles | 32 pages | **0** |
| Duplicate descriptions | — | **0** |
| Descriptions cut mid-word | 109 | **0** |
| Descriptions over 160 chars | 174 | **0** |
| Title length (incl. suffix) | up to "Overview" | **26–60** |
| Description length | 31–240 | **122–158** |

**No prose was rewritten.** I measured content overlap between the
colliding pairs first — only 17–23%, so these were always distinct pages
wearing the same label, not duplicate content. The `docs/` diff is **564
insertions, zero deletions**.

### Examples

```
BEFORE — all seven identical:  "Overview | Paperclip Docs"
AFTER:  "Adapters Overview"      "API Overview"        "CLI Overview"
        "Deployment Overview"    "Bundled Skills"      "Optional Skills"
        "Experimental Features"

BEFORE — both identical:  "Issues | Paperclip Docs"
AFTER:  "Issues: Creating and Tracking Work"   /guides/day-to-day/issues/
        "Issues API"                           /reference/api/issues/
```

```
/guides/getting-started/five-minute-path/
BEFORE (220):  The shortest route from a running Paperclip instance to an agent that has
               just completed work for you. After Paperclip is installed, the path takes
               about 5 minutes: create a company, hire a CEO agent, approve its first s
AFTER  (131):  Install Paperclip, create a company, hire a CEO agent, and approve its
               first strategy — a working AI company in about five minutes.
```

## Plumbing

- **`site/build-release.mjs`** prefers the authored fields; the derived
path stays only as a fallback the CI gate prevents ever reaching.
`<title>`, `<meta name="description">`, and the `TechArticle` JSON-LD
all read from the same source.
- **`site/app.js`** uses the same fields, so a client-side transition
cannot leave a different title on screen than the crawler was served.
Verified in-browser: navigating `/guides/day-to-day/issues/` →
`/reference/api/issues/` swaps to exactly the server-rendered values, no
JS errors.
- **`scripts/verify-seo-metadata.mjs`** (new) enforces presence,
uniqueness, length, complete-sentence endings, and the
frontmatter-safety rules. Wired into `docs:test` and CI as its own step.

I negative-tested the gate rather than assuming it works — it correctly
fails on a duplicate title, a mid-word description, and a missing field,
and passes again once restored.

## Automating it going forward

`skills/sync-docs/SKILL.md` now:

- gains an **"Authored SEO metadata"** section: why the fields exist,
the enforced rules as a table, and how to actually write them (title
names the subject not the nav position; description writes for the
click; never generate mechanically from the first paragraph — *that is
the fallback the gate exists to prevent*).
- requires both fields in the **page-authoring instruction** given to
every subagent, with guidance on when to update an existing description
versus leave it.
- runs `docs:test:seo-metadata` and `docs:test:crawlable-links` as
**fail-loud gates in Phase 7** before commit.
- documents all six build gates in the files/gates section.

So once this lands, a new page cannot ship without hand-written
metadata, and a title collision cannot reach production.

## Verified

All six verifiers pass:

```
Static route verification passed
Asset fingerprint verification passed
Hierarchical Skills nav verification passed
Skill source block verification passed (11 skill pages, 1 control page)
Crawlable-link contract passed: 40886 internal links across 194 pages resolve in one hop
Authored-SEO contract passed: 192 pages carry a unique, hand-written seo_title and seo_description
```

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude Opus 5 (1M context) <[email protected]>

2026-08-25 22:51:42 +02:00

7.5 KiB

Raw Permalink Blame History

paperclip_version, seo_title, seo_description

paperclip_version	seo_title	seo_description
v2026.609.0	The connect Command	The friendly front door for a fresh terminal — walks you through pointing the CLI at a running Paperclip instance in one guided flow.

Connect Command

When you sit down at a fresh terminal and want the Paperclip CLI to start talking to a running instance, connect is the friendly front door. It walks you through one short conversation — which server, which persona, which company and agent — and at the end it has minted a token, saved a reusable profile, and printed the environment exports you need. You do not have to know any IDs up front or hand-craft a context file; the wizard discovers what is available and lets you pick.

Reach for paperclipai connect the very first time you set up the CLI, whenever you want to add another saved profile (a second instance, a different company, a new agent), or any time you would rather click through choices than type out context set and auth login by hand.

Note: connect is interactive by design. It needs a real terminal (a TTY) on both input and output. In scripts and CI it will refuse to run and tell you to pass --api-base/--api-key or use the context and token commands instead. See Common Options for the non-interactive path.

What it does, end to end

connect is a wizard, not a group of subcommands. You run the single command and answer a handful of prompts:

paperclipai connect

Under the hood it walks through these steps:

Confirm the API base. It pre-fills the API base it would otherwise resolve (see Common Options for that resolution order) and lets you accept or change it. The placeholder is http://localhost:3100.
Health-check the server. It calls GET /api/health against the base you chose and stops early if the instance is not reachable — so you find connection problems now, not three commands later.
Log you in as a board operator. It runs the interactive board login to obtain a board credential, which it uses to discover the companies you can see.
Pick a persona. You choose whether this profile connects as a board operator or as an agent in a company.
Name the profile. It suggests the current profile name (or default) and lets you keep or rename it.
Choose company (and agent). A board profile can optionally pin a default company; an agent profile must pick a company and then an agent within it.
Mint a token, save the profile, print exports. It creates the right kind of token, writes a persona-aware profile into your context file, makes that profile current, and prints the shell exports you can paste to use the new credential immediately.

The two personas

The one decision that shapes everything else is persona. You can skip the prompt by passing --persona board or --persona agent; otherwise the wizard asks.

Persona	What it creates	Company & agent
`board`	A board API token, scoped to the instance.	Company is optional — you may pin a default for the profile or choose `(none)`. No agent.
`agent`	An agent API key, scoped to one company and one agent.	Company is required, and you then pick one of that company's agents. The company must have at least one agent.

A board profile carries board authority across the instance; an agent profile is bound to exactly one company and one agent. This is the same persona split described under Context profiles. For how each credential type is minted and what it can do, see Authentication.

# Connect straight as a board operator, skipping the persona prompt
paperclipai connect --persona board

# Connect as an agent (you'll still pick the company and agent interactively)
paperclipai connect --persona agent

Naming the token and the env var

Two flags let you control how the resulting credential is labelled and stored, so you do not have to accept the defaults.

paperclipai connect --token-name "laptop-board" --api-key-env-var-name PAPERCLIP_API_KEY

Flag	Default	Use
`--persona <persona>`	(prompted)	The persona to configure: `board` or `agent`. Pass it to skip the persona question.
`--api-key-env-var-name <name>`	`PAPERCLIP_API_KEY`	The name of the environment variable the profile will read its key from at call time. The profile stores this name, never the token itself.
`--token-name <name>`	A timestamped label	The label to give the created token. Defaults to `cli-board-<timestamp>` for a board token or `cli-agent-<timestamp>` for an agent key.

Tip: The profile records the name of the env var, not the secret. After connecting, keep the token in your environment (the printed exports do exactly this) so your context file stays safe to share. This is the same apiKeyEnvVarName mechanism explained in Common Options.

What you get back

When the wizard finishes it confirms the connection (Connected profile '<name>' as board. or ... as <agent name>.) and returns a result you can read or script against. Pass --json for the machine-readable form.

The result includes the profile name, the persona, the resolved apiBase, the companyId (and agentId/agentName for an agent profile), the created key (its id, name, createdAt, token, and — for a board token — expiresAt), and a ready-to-paste exports block:

export PAPERCLIP_API_URL='http://localhost:3100'
export PAPERCLIP_COMPANY_ID='<company-id>'
export PAPERCLIP_AGENT_ID='<agent-id>'
export PAPERCLIP_API_KEY='<token>'

The PAPERCLIP_COMPANY_ID and PAPERCLIP_AGENT_ID lines appear only when they apply, and the last line uses whatever name you set with --api-key-env-var-name. Paste the block into your shell and the rest of the CLI will pick up the new profile and credential immediately.

Warning: The plaintext token is shown once, in the wizard output and in the exports block. Capture it now — store it in your secret manager or export it into your environment. The saved profile only references the env var by name; it never holds the token for you to recover later.

Common Options

connect accepts the standard client options — --data-dir <path>, --api-base <url>, --api-key <token>, --context <path>, --profile <name>, -C, --company-id <id>, and --json. Of these, --api-base, --context, --profile, and --company-id seed the wizard's starting values (the API base it pre-fills, the context file it writes, the profile name it suggests, and the company it pre-selects). See Common Options for how API base, context, and company scope are resolved, and Output and Scripting for --json usage.

Note: connect performs an interactive board login as part of its flow, so passing --api-key is not the way to drive it non-interactively. If you need a scripted setup, mint a token another way and configure the profile with context set — the path connect itself points you to when there is no TTY.

7.5 KiB Raw Permalink Blame History