Strategy logo

Strategy

Community
galiprandi
strategy

Configure and adjust the job search strategy level. The agent interrogates the user, proposes a level, and saves it to DB. All flows respect it.

Overview

Publishergaliprandi
Repositoryjob-seeker
Skill namestrategy
Stars
68
Forks
11
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by galiprandi on GitHub. Read the source before you install it.

Installation

Install the Strategy AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/galiprandi/job-seeker/tree/main/.agents/skills/strategy
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/galiprandi/job-seeker.git /tmp/job-seeker
mkdir -p .claude/skills
cp -r /tmp/job-seeker/.agents/skills/strategy .claude/skills/strategy
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Strategy in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Strategy on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Strategy is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Strategy — Job search aggressiveness configuration

Trigger

Keyword: strategy

The user says strategy (or variants: "cambiar estrategia", "change strategy", "more aggressive", "less aggressive", "urgency", "how active") and the strategy configuration flow is triggered.

Purpose

The job search has configurable aggressiveness. Different situations (employed vs unemployed, urgent vs relaxed) require different levels of effort. This flow lets the user define and adjust their strategy, which all other flows respect.

Pre-flight

  • Load current strategy:
    bash
    node scripts/db.js "SELECT value FROM preferences WHERE user_id = <user_id> AND category = 'workflow' AND key = 'strategy_level' AND status = 'active'"
    node scripts/db.js "SELECT data->'strategy' AS strategy FROM users WHERE id = <user_id>"
  • If no strategy exists, note that onboarding step 4b was skipped. Default to selective.

Flow

1. Show current strategy

If a strategy exists, show the user:

Current strategy: selective
- Apply batch: 5 jobs per session
- Targets batch: 5 companies per session
- Daily frequency: 1x/day
- Match threshold: Must only
- Follow-up: 5 days
- Relax Must-haves: none
- Cold outreach: no
- Active sources: radar, apply, targets, referrals, news

2. Ask about current situation

Present the 4 levels (see AGENTS.md "Strategy levels") and ask:

  1. ¿Estás empleado actualmente?
  2. ¿Qué tan urgente es tu búsqueda?
    • Sin urgencia, solo mirando → passive
    • En los próximos meses, buscando algo mejor → selective
    • Necesito algo pronto → active
    • Necesito algo ya, desesperado → aggressive
  3. Pregunta adaptativa según career stage (leer users.data.profile.career_stage y users.data.profile.has_management):
    • Si career_stage es junior/mid: "¿Estás abierto a roles de un nivel más alto al tuyo, o solo a roles de tu mismo nivel?"
    • Si career_stage es senior+ y has_management = true: "¿Aceptarías roles IC o solo Manager?"
    • Si career_stage es senior+ y has_management = false: "¿Te interesa dar el salto a management o prefieres seguir como IC?"
  4. ¿Aceptarías hybrid si el proyecto es muy bueno?
  5. ¿Quieres que aplique automáticamente o solo te muestre opciones?

Nunca preguntes "IC o Manager?" a un usuario junior/mid. La pregunta 3 se adapta al perfil inferido del CV.

Based on answers, propose a level. Explain what changes:

Based on your answers, I propose: active

This means:
- I'll apply to 10 jobs per session (Must + Strong matches)
- Register on 10 target companies per session
- Run daily 2x/day
- Follow up after 3 days instead of 5
- Relax your top 2 Must-haves (the agent resolves which ones from your profile)
- Send cold outreach to recruiters at target companies

Does this work? You can adjust any parameter individually.

3. Allow customization

After proposing a level, let the user customize individual parameters:

  • "Quiero que apliques a 15 pero solo Must-matches" → override apply_batch_size = 15, keep match_threshold = must_only
  • "No quiero cold outreach" → override cold_outreach = false
  • "Relajar remote pero no mentorship" → override relax_must_haves with specific keys from the user's Must-haves (the agent reads users.data.job_preferences to resolve which keys to relax)

4. Save to DB

bash
# Save level
node scripts/db.js "INSERT INTO preferences (user_id, category, key, value, confidence, source) VALUES (<user_id>, 'workflow', 'strategy_level', '<level>', 1.0, 'explicit_statement') ON CONFLICT (user_id, category, key) DO UPDATE SET value = EXCLUDED.value, source = EXCLUDED.source, updated_at = NOW()" --write

# Save detailed parameters
node scripts/db.js "UPDATE users SET data = jsonb_set(data, '{strategy}', '<json>'::jsonb) WHERE id = <user_id>" --write

The strategy JSON contains all parameters (see AGENTS.md "Strategy levels"). Example for active with customizations:

json
{
  "level": "active",
  "apply_batch_size": 15,
  "targets_batch_size": 10,
  "daily_frequency": "2x/day",
  "match_threshold": "must_strong",
  "follow_up_days": 3,
  "relax_must_haves": "top_2_must_haves",
  "relaxed_keys": ["remote", "salary"],
  "cold_outreach": false,
  "sources_active": ["radar", "apply", "targets", "referrals", "news"]
}

relax_must_haves stores the strategy level (top_2_must_haves, top_3_must_haves, none). relaxed_keys is the resolved list of actual Must-have keys from users.data.job_preferences that the agent will relax. The agent populates relaxed_keys at runtime by reading the user's Must-weighted preferences and picking the top N by priority.

5. Confirm

Show the final strategy and confirm:

Strategy saved: active (customized)

- Apply: 15 jobs/session, Must+Strong matches
- Targets: 10 companies/session
- Daily: 2x/day
- Follow-up: 3 days
- Relaxed: top 2 Must-haves (remote, salary)
- Cold outreach: disabled
- Sources: radar, apply, targets, referrals, news

All flows will respect this. Say "strategy" again to change it.

Level defaults

Canonical level table (this is the source of truth):

LevelSituationapply_batchtargets_batchdailymatchfollow_uprelaxcoldsources
passiveEmployed, open to opportunities00on-demandmust_only7nonefalseradar, news
selectiveEmployed, looking for better551x/daymust_only5nonefalseradar, apply, targets, referrals, news
activeUnemployed or about to be10102x/daymust_strong3top_2_must_havestrueradar, apply, targets, referrals, news
aggressiveNeeds a job now15all2x/daymust_strong_nice2top_3_must_havestrueradar, apply, targets, referrals, news

relax_must_haves (none/top_2_must_haves/top_3_must_haves) resolves to actual Must-have keys from users.data.job_preferences at runtime — never assume which Must-haves exist. sources_active controls which pillars each flow runs; absent source → skipped entirely. apply_batch_size = 0 → present matches for manual approval only.

Rules

  • Always show current strategy first before proposing changes
  • Explain what changes when proposing a new level. The user needs to understand the impact
  • Allow customization of any parameter after choosing a level. Don't force all defaults
  • Save both the level name (preferences) and detailed params (users.data.strategy). The level is the quick reference, the params are what flows actually read
  • Report what was saved (Gold Rule 3). One line: "Estrategia actualizada: active"
  • Never change strategy without asking. Even if memory detects a situation change, propose the change and wait for confirmation
  • Single user (repo owner)

Dependencies

  • Depends on onboarding (DB must exist)
  • Read by all flows at pre-flight
  • Memory skill can trigger this flow when situation changes are detected

Timing and batch sizes

Practical limits per session (validated empirically):

  • An apply session can process 7-10 Easy Apply jobs in ~30 min
  • Connection requests: 8-10 per session (avoid LinkedIn limits)
  • Direct emails: 4-5 per session (each takes ~2 min with attachment)
  • Effective total per session: 15-20 application/contact actions
  • Some companies have very long forms that take ~10 min each. The rest take 2-5 min each

These limits interact with strategy levels:

  • passive: 0 applications, 0 targets (on-demand only)
  • selective: 5 applications, 5 targets per session
  • active: 10 applications, 10 targets per session
  • aggressive: 15 applications, all targets per session

LinkedIn limits: custom notes on connection requests have a weekly cap. When exhausted, send invites without a note. Don't retry with a note.

Frequently asked questions

What does the Strategy AI skill do?

Configure and adjust the job search strategy level. The agent interrogates the user, proposes a level, and saves it to DB. All flows respect it.

Why use Strategy on TypingMind?

Because you install it once and use it with any model. Strategy is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Strategy in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/galiprandi/job-seeker/tree/main/.agents/skills/strategy. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Strategy?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Strategy?

As many as you like. As long as a model supports skills, you can use Strategy with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Strategy AI skill free?

Yes. It is published on GitHub by galiprandi under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇