hyperhive/hive-c0re/src/forge/users.rs

//! The bootstrap `core` admin user: its account, access-token lifecycle
//! and avatar upload. The typed API-client constructor + `forgejo admin`
//! helpers live in the module root (`super`).

use std::path::Path;

use anyhow::{Context, Result};
use base64::Engine;
use forgejo_api::structs::UpdateUserAvatarOption;
use forgejo_api::{ApiErrorKind, ForgejoError};
use reqwest::StatusCode;

use super::{api, bounded, forge_admin};

const TOKEN_NAME_PREFIX: &str = "hyperhive";
// Where the host-side `core` admin token lives. Used by hive-c0re itself
// to push the meta repo + drive admin API calls. Root-only. Literal lives
// in `crate::paths`; aliased here under the long-standing name.
use crate::paths::FORGE_CORE_TOKEN as CORE_TOKEN_PATH;
// The `forge/core-avatar-set` marker lives in `crate::paths` — a one-shot
// guard: the upload runs once, the marker is written, subsequent startups
// skip. Delete it to force a re-upload.
// The avatar PNG path is resolved by `core_avatar_png_path` below —
// only-used-here, so it lives in this module rather than the shared
// `hive_sh4re::assets` helpers (which now hold only `prompt_template`, the
// one asset path every crate needs — the agent icon, the last other one, moved off the
// shared-asset model entirely; see `hive-agent::web_ui::screen`).

/// `$HIVE_ASSETS_DIR/branding/hyperhive.png` — the core mark,
/// rasterised. Not independently configurable — override the whole
/// `services.hyperhive.c0re.assets` package to change it.
fn core_avatar_png_path() -> std::path::PathBuf {
    let dir = std::env::var("HIVE_ASSETS_DIR").expect(
        "HIVE_ASSETS_DIR is unset — the hyperhive NixOS module sets it from \
         services.hyperhive.c0re.assets on the hive-c0re unit, so this \
         process was started outside that unit",
    );
    std::path::PathBuf::from(dir).join("branding/hyperhive.png")
}

/// Bootstrap `core` token scopes — the same set as swarm-controller's
/// `AGENT_TOKEN_SCOPES`, with no admin scope: nothing in hive-c0re calls
/// `/api/v1/admin/*`. See `docs/integrations/forge.md::Token scopes`.
const CORE_TOKEN_SCOPES: &str = "read:user,write:user,read:notification,write:notification,write:repository,write:issue,write:organization,write:misc";

/// Pull the access token out of forgejo's success message. Format
/// has shifted across versions (table form vs. "Access token was
/// successfully created: `<hex>`"), so just hunt the output for the
/// first long hex-looking word.
fn extract_token(output: &str) -> Option<String> {
    output
        .split(|c: char| c.is_whitespace() || c == ',' || c == ':')
        .find(|w| w.len() >= 32 && w.chars().all(|c| c.is_ascii_hexdigit()))
        .map(str::to_owned)
}

/// Canonical email address for a hive agent's Forgejo account.
/// Must match the `user.email` set by `meta::render_flake` so commits
/// by the agent link back to their Forgejo profile page.
fn agent_email(name: &str) -> String {
    format!("{name}@hyperhive.local")
}

/// Ensure a forgejo user named `name` exists. Idempotent: forgejo
/// returns a "user already exists" error which we treat as success.
/// `admin` adds `--admin` (site admin) — used for the bootstrap
/// `core` user that drives the API. The password is random and never
/// read: the account authenticates by token.
async fn ensure_user_exists(name: &str, admin: bool) -> Result<()> {
    let email = agent_email(name);
    let mut args = vec![
        "user",
        "create",
        "--username",
        name,
        "--email",
        &email,
        "--random-password",
        "--must-change-password=false",
    ];
    if admin {
        args.push("--admin");
    }
    let result = forge_admin(&args).await;
    match result {
        Ok(_) => {
            tracing::info!(%name, "forge: created user");
            Ok(())
        }
        Err(e) => {
            // Forgejo's "already exists" error wording varies; just
            // try the next step and let token issuance surface a
            // real failure if the user truly isn't there.
            let msg = format!("{e:#}");
            if msg.contains("already exists") || msg.contains("user already") {
                tracing::debug!(%name, "forge: user already exists");
                Ok(())
            } else {
                tracing::warn!(%name, error = %msg, "forge: user create unclear; trying token anyway");
                Ok(())
            }
        }
    }
}

/// Mint a fresh access token for `name`. Token name is suffixed with
/// a monotonic clock so re-issuing doesn't collide with an existing
/// token of the same name in the DB. `scopes` is the scope string
/// passed to `forgejo admin user generate-access-token --scopes`;
/// the bootstrap `core` user, its one caller, passes `CORE_TOKEN_SCOPES`.
async fn mint_token(name: &str, scopes: &str) -> Result<String> {
    let token_name = format!(
        "{TOKEN_NAME_PREFIX}-{}",
        std::time::SystemTime::now()
            .duration_since(std::time::UNIX_EPOCH)
            .map_or(0, |d| d.as_secs())
    );
    let stdout = forge_admin(&[
        "user",
        "generate-access-token",
        "--username",
        name,
        "--token-name",
        &token_name,
        "--scopes",
        scopes,
    ])
    .await?;
    // Deliberately does NOT include `stdout`: on this path forgejo has
    // already minted a live token and printed it, and an error string
    // propagates into logs the same way any other message does. The
    // shape of the output is enough to diagnose a version drift.
    let token = extract_token(&stdout).with_context(|| {
        format!(
            "no token-shaped word (>= 32 hex chars) in forgejo output \
             ({} bytes, {} lines); output withheld, it carries the token",
            stdout.len(),
            stdout.lines().count()
        )
    })?;
    tracing::debug!(%name, %token_name, "forge: minted access token");
    Ok(token)
}

/// Mint a fresh Forgejo access token for the `core` admin user and
/// write it directly to `path`. Unlike agent tokens this path is owned
/// by hive-c0re itself (under `/var/lib/hyperhive/`), so a direct
/// write is both correct and necessary (no priv round-trip).
async fn mint_and_persist_core_token(path: &Path) -> Result<()> {
    let token = mint_token("core", CORE_TOKEN_SCOPES).await?;
    crate::secret_file::write_secret_0600(path, format!("{token}\n").as_bytes())
        .with_context(|| format!("write core token to {}", path.display()))?;
    tracing::info!(path = %path.display(), "forge: persisted core access token");
    Ok(())
}

/// Set `core`'s Forgejo avatar to the hyperhive logo once, then
/// remember it so subsequent startups don't re-upload. Best-effort
/// — any non-2xx is logged at the caller; the project runs fine
/// with the default hash identicon.
pub(super) async fn ensure_core_avatar(token: &str) -> Result<()> {
    let marker = crate::paths::forge_core_avatar_marker();
    if marker.exists() {
        return Ok(());
    }
    let png_path = core_avatar_png_path();
    let png_bytes = tokio::fs::read(&png_path)
        .await
        .with_context(|| format!("read core avatar PNG from {}", png_path.display()))?;
    // The raw-HTTP predecessor POSTed the admin endpoint
    // `/admin/users/core/avatar`, which forgejo-api has no method for.
    // `token` IS the core user's own token though, so updating "the
    // current user's avatar" (`POST /user/avatar`) is behaviorally
    // identical.
    bounded(
        "set core avatar",
        api(token)?.user_update_avatar(UpdateUserAvatarOption {
            image: Some(base64::engine::general_purpose::STANDARD.encode(&png_bytes)),
        }),
    )
    .await?
    .context("set core avatar")?;
    if let Err(e) = write_marker(&marker) {
        tracing::warn!(marker = %marker.display(), error = ?e, "forge: could not write the avatar marker; the upload repeats on the next start");
    }
    tracing::info!("forge: set core user avatar to hyperhive logo");
    Ok(())
}

/// Create the one-shot marker file, and its directory if needed.
fn write_marker(marker: &Path) -> std::io::Result<()> {
    if let Some(parent) = marker.parent() {
        std::fs::create_dir_all(parent)?;
    }
    std::fs::write(marker, "")
}

/// Outcome of probing whether the persisted core token still works
/// against the *current* forge. Existence on disk is not validity: a
/// token minted before a forge rebuild / re-provision is unknown to the
/// new forge's DB and 401s on every call.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
enum CoreTokenCheck {
    /// Token authenticated successfully — keep using it.
    Valid,
    /// Forge explicitly rejected the token (401/403) — re-mint.
    Invalid,
    /// Couldn't determine (forge unreachable / 5xx). Don't re-mint on a
    /// transient: keep the existing token and let a later ensure pass
    /// re-check once the forge is responsive. Re-minting here would both
    /// fail (mint needs the forge too) and churn tokens needlessly.
    Indeterminate,
}

/// Map a failed token-probe call to a [`CoreTokenCheck`]. Only a
/// definitive auth rejection (401/403 — surfaced by the typed client
/// as `Unauthorized`/`Forbidden` API errors, or defensively as bare
/// `UnexpectedStatusCode`s) is `Invalid`; anything else (transport,
/// 5xx, unexpected shapes) is `Indeterminate`. Pure so the decision
/// logic is unit-testable without a live forge.
fn classify_core_token_error(e: &ForgejoError) -> CoreTokenCheck {
    match e {
        ForgejoError::ApiError(api) => match api.error_kind() {
            ApiErrorKind::Unauthorized | ApiErrorKind::Forbidden => CoreTokenCheck::Invalid,
            _ => CoreTokenCheck::Indeterminate,
        },
        ForgejoError::UnexpectedStatusCode(s)
            if *s == StatusCode::UNAUTHORIZED || *s == StatusCode::FORBIDDEN =>
        {
            CoreTokenCheck::Invalid
        }
        _ => CoreTokenCheck::Indeterminate,
    }
}

/// Probe whether `token` is still accepted by the current forge with a
/// cheap authenticated `user_get_current` (`GET /api/v1/user`, covered
/// by the core token's `read:user` scope). See [`CoreTokenCheck`] for
/// how the outcome is interpreted.
async fn check_core_token(token: &str) -> CoreTokenCheck {
    let Ok(client) = api(token) else {
        return CoreTokenCheck::Indeterminate;
    };
    let Ok(probe) = bounded("probe core token", client.user_get_current()).await else {
        tracing::debug!("forge: core-token probe timed out; treating as indeterminate");
        return CoreTokenCheck::Indeterminate;
    };
    match probe {
        Ok(_) => CoreTokenCheck::Valid,
        Err(e) => {
            let outcome = classify_core_token_error(&e);
            if outcome == CoreTokenCheck::Indeterminate {
                tracing::debug!(
                    error = %e,
                    "forge: core-token probe inconclusive (unreachable / unexpected response); \
                     treating as indeterminate"
                );
            }
            outcome
        }
    }
}

/// Ensure the bootstrap `core` admin user + a token at
/// `CORE_TOKEN_PATH`. The token is what hive-c0re uses for forgejo
/// API calls (meta-repo push, collaborator grants).
/// Returns the token.
///
/// Idempotent, but validity-aware: when a token file is already present
/// it is **probed against the current forge** before being trusted. A
/// token persisted before a forge rebuild / re-provision is stale (the
/// new forge DB doesn't know it) and would 401 every caller — so on a
/// definitive rejection the token is re-minted. A merely-unreachable
/// forge leaves the existing token in place (a later ensure pass
/// re-checks) rather than churning tokens on a transient.
pub(super) async fn ensure_core_user_and_token() -> Result<String> {
    let path = std::path::Path::new(CORE_TOKEN_PATH);
    if let Ok(existing) = std::fs::read_to_string(path) {
        let trimmed = existing.trim().to_owned();
        if !trimmed.is_empty() {
            match check_core_token(&trimmed).await {
                CoreTokenCheck::Valid | CoreTokenCheck::Indeterminate => return Ok(trimmed),
                CoreTokenCheck::Invalid => {
                    tracing::warn!(
                        path = %path.display(),
                        "forge: persisted core token rejected by forge (stale after rebuild?); \
                         re-minting"
                    );
                }
            }
        }
    }
    ensure_user_exists("core", true).await?;
    mint_and_persist_core_token(path).await?;
    let raw = std::fs::read_to_string(path)
        .with_context(|| format!("read {CORE_TOKEN_PATH} after mint"))?;
    Ok(raw.trim().to_owned())
}

/// Read the persisted core token, or None when the forge isn't
/// seeded yet. Cheap — just a file read.
pub fn core_token() -> Option<String> {
    std::fs::read_to_string(CORE_TOKEN_PATH)
        .ok()
        .map(|s| s.trim().to_owned())
        .filter(|s| !s.is_empty())
}

#[cfg(test)]
mod tests {
    use super::{CoreTokenCheck, classify_core_token_error, write_marker};
    use forgejo_api::{ApiError, ApiErrorKind, ForgejoError};
    use reqwest::StatusCode;

    #[test]
    fn write_marker_creates_the_missing_parent_dir() {
        let dir = tempfile::tempdir().expect("tempdir");
        let marker = dir.path().join("not-yet").join("nested").join("avatar-set");
        write_marker(&marker).expect("write");
        assert!(marker.is_file());
    }

    /// A parent that is a regular file makes `create_dir_all` fail for any
    /// uid, root included.
    #[test]
    fn write_marker_reports_a_path_it_cannot_create() {
        let dir = tempfile::tempdir().expect("tempdir");
        let file = dir.path().join("a-file");
        std::fs::write(&file, "").expect("file");
        assert!(write_marker(&file.join("sub").join("avatar-set")).is_err());
    }

    fn api_err(kind: ApiErrorKind) -> ForgejoError {
        ForgejoError::ApiError(ApiError {
            message: None,
            kind,
        })
    }

    #[test]
    fn auth_rejection_errors_are_invalid() {
        // The whole point: a stale token (forge rebuilt out from under it)
        // 401s, and 401/403 are the only outcomes that trigger a re-mint.
        assert_eq!(
            classify_core_token_error(&api_err(ApiErrorKind::Unauthorized)),
            CoreTokenCheck::Invalid
        );
        assert_eq!(
            classify_core_token_error(&api_err(ApiErrorKind::Forbidden)),
            CoreTokenCheck::Invalid
        );
        // Defensive: the same statuses arriving as bare status codes
        // (endpoint spec didn't list them) must classify identically.
        for s in [StatusCode::UNAUTHORIZED, StatusCode::FORBIDDEN] {
            assert_eq!(
                classify_core_token_error(&ForgejoError::UnexpectedStatusCode(s)),
                CoreTokenCheck::Invalid,
                "status {s} should be invalid"
            );
        }
    }

    #[test]
    fn transient_and_unexpected_errors_are_indeterminate() {
        // Never re-mint on a transient — minting needs the forge too, and
        // churning tokens on a blip is worse than keeping the existing one.
        for s in [
            StatusCode::INTERNAL_SERVER_ERROR,
            StatusCode::BAD_GATEWAY,
            StatusCode::SERVICE_UNAVAILABLE,
            StatusCode::GATEWAY_TIMEOUT,
            StatusCode::NOT_FOUND,
        ] {
            assert_eq!(
                classify_core_token_error(&ForgejoError::UnexpectedStatusCode(s)),
                CoreTokenCheck::Indeterminate,
                "status {s} should be indeterminate"
            );
        }
        assert_eq!(
            classify_core_token_error(&api_err(ApiErrorKind::NotFound { errors: None })),
            CoreTokenCheck::Indeterminate
        );
        assert_eq!(
            classify_core_token_error(&api_err(ApiErrorKind::Generic)),
            CoreTokenCheck::Indeterminate
        );
    }
}